Large-scale molecular phylogeny, morphology, divergence-time estimation, and the fossil record of advanced caenophidian snakes (Squamata: Serpentes)

PLOS ONE, May 2019

Caenophidian snakes include the file snake genus Acrochordus and advanced colubroidean snakes that radiated mainly during the Neogene. Although caenophidian snakes are a well-supported clade, their inferred affinities, based either on molecular or morphological data, remain poorly known or controversial. Here, we provide an expanded molecular phylogenetic analysis of Caenophidia and use three non-parametric measures of support–Shimodaira-Hasegawa-Like test (SHL), Felsentein (FBP) and transfer (TBE) bootstrap measures–to evaluate the robustness of each clade in the molecular tree. That very different alternative support values are common suggests that results based on only one support value should be viewed with caution. Using a scheme to combine support values, we find 20.9% of the 1265 clades comprising the inferred caenophidian tree are unambiguously supported by both SHL and FBP values, while almost 37% are unsupported or ambiguously supported, revealing the substantial extent of phylogenetic problems within Caenophidia. Combined FBP/TBE support values show similar results, while SHL/TBE result in slightly higher combined values. We consider key morphological attributes of colubroidean cranial, vertebral and hemipenial anatomy and provide additional morphological evidence supporting the clades Colubroides, Colubriformes, and Endoglyptodonta. We review and revise the relevant caenophidian fossil record and provide a time-calibrated tree derived from our molecular data to discuss the main cladogenetic events that resulted in present-day patterns of caenophidian diversification. Our results suggest that all extant families of Colubroidea and Elapoidea composing the present-day endoglyptodont fauna originated rapidly within the early Oligocene–between approximately 33 and 28 Mya–following the major terrestrial faunal turnover known as the “Grande Coupure” and associated with the overall climate shift at the Eocene-Oligocene boundary. Our results further suggest that the caenophidian radiation originated within the Caenozoic, with the divergence between Colubroides and Acrochordidae occurring in the early Eocene, at ~ 56 Mya.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0216148&type=printable

Large-scale molecular phylogeny, morphology, divergence-time estimation, and the fossil record of advanced caenophidian snakes (Squamata: Serpentes)

May Large-scale molecular phylogeny, morphology, divergence-time estimation, and the fossil record of advanced caenophidian snakes (Squamata: Serpentes) Hussam ZaherID 0 2 Robert W. Murphy 1 2 Juan Camilo Arredondo 0 2 Roberta Graboski 0 2 Paulo Roberto Machado-Filho 0 2 Kristin Mahlow 2 Giovanna G. Montingelli 0 2 Ana Bottallo Quadros 0 2 Nikolai L. Orlov 2 Mark Wilkinson 2 Ya-Ping Zhang 2 Felipe G. Grazziotin 2 0 Museu de Zoologia, Universidade de Sa?o Paulo , Sa?o Paulo, Sa?o Paulo, Brazil, 2 CR2P - Centre de Recherche en Pale ? ontologie - Mus e ?um national d'Histoire naturelle - Sorbonne Universite ? , Paris , France 1 Centre for Biodiversity, Royal Ontario Museum , Toronto, Ontario , Canada , 4 State Key Laboratory of Genetic Resources and Evolution, Kunming Institute of Zoology , Kunming, China, 5 Laborato ?rio de Herpetologia, Museu Paraense Em ??lio Goeldi, Bele ? m, Par a ?, Brazil, 6 Museum fu ? r Naturkunde , Leibniz Institute for Evolution and Biodiversity Science , Berlin, Germany , 7 Zoological Institute, Russian Academy of Sciences , Saint Petersburg , Russia , 8 Department of Life Sciences , The Natural History Museum, London , United Kingdom , 9 Laboratory for Conservation and Utilization of Bio-resources, Yunnan University , Kunming, China, 10 Laborat o ?rio de Colec ?o?es Zool o ?gicas , Instituto Butantan , Sa?o Paulo, Sa?o Paulo , Brazil 2 Editor: Ulrich Joger , State Museum of Natural History , GERMANY Caenophidian snakes include the file snake genus Acrochordus and advanced colubroidean snakes that radiated mainly during the Neogene. Although caenophidian snakes are a wellsupported clade, their inferred affinities, based either on molecular or morphological data, remain poorly known or controversial. Here, we provide an expanded molecular phylogenetic analysis of Caenophidia and use three non-parametric measures of support-Shimodaira-Hasegawa-Like test (SHL), Felsentein (FBP) and transfer (TBE) bootstrap measuresto evaluate the robustness of each clade in the molecular tree. That very different alternative support values are common suggests that results based on only one support value should be viewed with caution. Using a scheme to combine support values, we find 20.9% of the 1265 clades comprising the inferred caenophidian tree are unambiguously supported by both SHL and FBP values, while almost 37% are unsupported or ambiguously supported, revealing the substantial extent of phylogenetic problems within Caenophidia. Combined FBP/TBE support values show similar results, while SHL/TBE result in slightly higher combined values. We consider key morphological attributes of colubroidean cranial, vertebral and hemipenial anatomy and provide additional morphological evidence supporting the clades Colubroides, Colubriformes, and Endoglyptodonta. We review and revise the relevant caenophidian fossil record and provide a time-calibrated tree derived from our molecular data to discuss the main cladogenetic events that resulted in present-day patterns of - Data Availability Statement: All relevant data are within the paper and its Supporting Information files. caenophidian diversification. Our results suggest that all extant families of Colubroidea and Elapoidea composing the present-day endoglyptodont fauna originated rapidly within the was provided by Fundac??o de Amparo ? Pesquisa do Estado de S?o Paulo (BIOTA/FAPESP grants 2002/13602-4 and 2011/50206-9 to HZ and 2016/ 50127-5). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. early Oligocene?between approximately 33 and 28 Mya?following the major terrestrial faunal turnover known as the ?Grande Coupure? and associated with the overall climate shift at the Eocene-Oligocene boundary. Our results further suggest that the caenophidian radiation originated within the Caenozoic, with the divergence between Colubroides and Acrochordidae occurring in the early Eocene, at ~ 56 Mya. Introduction Determining the phylogenetic affinities within snakes was viewed by many herpetologists in the past as an insurmountable challenge. Underwood [1] expressed his profound frustration with a simple sentence: "I have found snake systematics to be a hard test to intellectual honesty?. Although the phylogenetic affinities of snakes were indeed difficult to determine on morphological grounds, monophyly of some higher-level taxa represent a long-standing consensus. This is the case for the clade Caenophidia, a group of advanced alethinophidian snakes recognized formally by Hoffstetter [ 2 ] to accommodate the families Colubridae, Dipsadidae, Hydrophiidae, Elapidae, and Viperidae. Hoffstetter?s Caenophidia was characterized by the absence of a coronoid bone and included the colubrid subfamily Acrochordinae, already known to share several additional derived morphological traits with the remaining caenophidian families [ 3,4 ]. The same group of ?advanced alethinophidian snakes? was also recognized by Romer [ 5 ], who preferred to accommodate them in a newly erected superfamily Colubroidea, equating the latter with Hoffstetter?s concept of Caenophidia. ?Acrochordoids? and ?colubroids? were only later recognized as two distinct superfamilies within Caenophidia after Groombridge [ 6,7 ] argued convincingly that acrochordids were the sister-group of the remaining caenophidians based on a number of synapomorphies derived from the vomeronasal capsule, musculature, hyoid and costal cartilages [ 8,9,10 ]. Molecular phylogenies ultimately provided strong support for the monophyly of Caenophidia, and further corroborated more controversial morphological hypotheses, such as the polyphyly of solenoglyphous [ 11 ] and proteroglyphous snakes [ 12 ]. On the other hand, analyses of molecular evidence also obtained conflicting results for the positions of acrochordids and xenodermids at the base of the Caenophidian tree and highlighted the need of substantial taxonomic changes in order to obtain monophyletic familial level taxa [ 13?28 ]. Thus, despite notable advances, many questions regarding the higher-level phylogeny and taxonomy of Caenophidia remain unanswered, and a period of taxonomic instability has seen a number of different, and sometimes contradictory, classification schemes, with none of them being entirely satisfactory (S1 Table). Three large-scale molecular phylogenies of snakes were published recently [ 26,27,28 ]. However, despite their impressive taxon sampling, substantial overlap in data and similar analytical strategies, these studies have produced a surprisingly large number of differences in inferred relationships at the familial and generic levels (Figs 1?3). Pyron et al. [ 26 ] and Figueroa et al. [ 28 ] based a number of taxonomic actions exclusively on their molecular phylogenetic analyses with no attempt to reconcile these with the available morphological and paleontological evidence. This is understandable given that one of the main advantages of molecular over morphological phylogenetics is the wider coverage of species that the technique allows within a relatively short amount of time. Large-sample comparative morphological studies are often difficult to achieve due to the need for destructive investigative procedures on limited museum specimens and by the lack of 2 / 82 Fig 1. Higher-level caenophidian phylogenies. Comparison between Maximum-likelihood phylogenetic estimates from (A) the present study and (B) Figueroa et al. [ 28 ]. Tips represent commonly recognized families, subfamilies and rogue taxa. Names in red correspond to taxa with distinct phylogenetic positions in the topologies compared. Numbers on each branch and within expanded tips correspond to our and previously reported support values: (A) FBP (left) and SHL (right); (B) FBP. Branches without numbers have support <70%. 3 / 82 Fig 2. Higher-level caenophidian phylogenies. Comparison between Maximum-likelihood phylogenetic estimates from (A) the present study and (B) Pyron et al. [ 26 ]. Tips represent commonly recognized families, subfamilies and rogue taxa. Names in red correspond to taxa with distinct phylogenetic positions in the topologies compared. Numbers on each branch and within expanded tips correspond to our and previously reported support values: (A) FBP (left) and SHL (right); (B) SHL. Branches without numbers have support <70%. 4 / 82 Fig 3. Higher-level caenophidian phylogenies. Comparison between Maximum-likelihood phylogenetic estimates from (A) the present study and (B) Zheng and Wiens [ 27 ]. Tips represent commonly recognized families, subfamilies and rogue taxa. Names in red correspond to taxa with distinct phylogenetic positions in the topologies compared. Numbers on each branch and within expanded tips correspond to our and previously reported support values: (A) FBP (left) and SHL (right); (B) FBP. Branches without numbers have support <70%. 5 / 82 comprehensive taxonomic coverage in skeletal collections all over the world. Morphological information on caenophidian snake anatomy is still very limited compared to the diversity within the group, with only a few detailed analyses of larger clades being available in the literature and mainly focused on the cranial and hemipenial complexes [ 11,29 ]. A notable exception is the monographic study of Cundall and Irish [30] of the skull of snakes, which provides the first comprehensive large-scale comparative analysis of caenophidian cranial anatomy. The paleontological record for caenophidian snakes is largely biased towards disarticulated postcranial (vertebral) materials [ 31,32 ]. Although caenophidian vertebral elements are frequently found in Caenozoic vertebrate-bearing deposits, their identification at the generic and familial levels are often difficult to ascertain, mostly because of our limited knowledge of vertebral morphology and its variation within caenophidian families. However, although limited, our present knowledge on the cranial, vertebral, and hemipenial anatomy of the group still constitutes an important body of evidence that can be evaluated within an explicit molecular phylogenetic framework, helping highlight major events in the origin and diversification of caenophidian snakes. This approach can help circumvent conflicts between multiple alternative molecular hypotheses of relationships [ 26,27,28 ] that seem to be correlated with poorly sampled groups or short internal branches combined with terminal taxa with long branches resulting from the accumulation of molecular autapomorphies [21]. Here, we provide an expanded molecular phylogenetic tree of Caenophidia, highlighting strongly and weakly supported hypotheses of relationships that need further investigation. We evaluate alternatively three non-parametric support values?Shimodaira-Hasegawa-Like test (SHL), Felsentein bootstrap proportions (FBP), and transfer bootstrap expectation metrics (TBE)?and combine two of these (FBP and SHL) to give a seven-category classification of the robustness of clades in the molecular tree (S2 Table). Contradictory support values are frequently encountered, suggesting that results based on only one of these three support values, should be viewed with caution. We use our molecular tree as a backbone phylogeny to review some key morphological characters of caenophidian snakes in an attempt to reconcile both morphological and molecular bodies of evidence at the familial and suparfamilial levels of the tree. We focus on two main anatomical complexes in the skull of caenophidian snakes?the optic nerve foramen/fenestra and the naso-frontal joint?known to be phylogenetically informative at higher levels [ 30 ]. We also revise anatomical evidence from the vertebrae and hemipenes of representatives of all known extant colubroidean families. Finally, we combine information from the known fossil record and a time-calibrated tree derived from our molecular data to discuss the main cladogenetic events that resulted in present-day patterns of caenophidian diversification. Materials and methods Taxonomic background Lawson et al. [ 18 ], Zaher et al. [ 23 ], and Pyron et al. [ 26 ] provided a listing of extant genera considered valid under their family-group names, while Uetz et al. [ 33 ] and Wallach et al. [34] went further and compiled complete listings of all extant species. In addition to known extant taxa, Wallach et al. [34] also provided a listing of extinct genera and species. Uetz et al. [ 33 ] species list represents a compilation of valid names that follows, in most respects, the latest taxonomic opinions, and thus can be highly unstable. On the other hand, Wallach et al.?s [34] work includes a large number of changes and corrections based on their own critical taxonomic opinion. In that sense, the latter work is a valuable source of original information. However, Uetz et al. [ 33 ] taxonomic list seems to integrate more accurately the massive contribution of the Herpetological community in recent years and we use that as a framework to 6 / 82 describe our results with respect to valid genera and species. However, we consider both taxonomic schemes at the family level to be problematic in several respects and follow instead the supra-familial and familial taxonomic scheme proposed by Zaher et al. [ 23,25 ], expanded here to include recently erected or recognized Pseudoxyrhophiinae, Grayiinae, Prosymnidae, and Pseudaspididae [ 22,26 ]. We also followed Savage [35] in the use of the family names Pareidae and Xenodermidae instead of Pareatidae and Xenodermatidae. Recently, Weinell and Brown [36] resolved the phylogenetic affinities of the genera Cyclocorus and Oxyrhabdium, which along with Hologerrhum and Myersophis, were retrieved in their analysis as a well-supported elapoid clade of endemic Philippine snakes. Weinell and Brown [36] accommodated these four genera in a new subfamily, Cyclocorinae, further referred herein as a family, Cyclocoridae. As a result, we considered the following 19 families here: Xenodermidae, Pareidae, Viperidae, Homalopsidae, Elapidae, Psammophiidae, Atractaspididae, Pseudoxyrhophiidae, Lamprophiidae, Cyclocoridae, Prosymnidae, Pseudaspididae, Sibynophiidae, Calamariidae, Grayiidae, Colubridae, Pseudoxenodontidae, Dipsadidae, and Natricidae. The potential taxonomic instability generated by unstable species or species groups representing rogue taxa [37] in molecular analyses is also of concern here. Recently, many of these taxa were given new generic or subgeneric names by R. Hoser who?s approach is considered unethical and potentially harmful for taxonomic stability, resulting in a request by a large consorsium of herpetologists, that the International Commission of Zoological Nomenclature (ICZN) invalidate these new names [38], a petition we strongly support. We refrain from using these names until a definitive decision on the validity of such names is reached by the ICZN. Taxon and gene sampling We assembled a data matrix comprising 1278 (15 outgroup and 1263 ingroup) terminal taxa representing all caenophidian families (see S3 Table for number of genes and accession numbers; see also S4 and S5 Tables for more details on taxon sampling). We obtained 5063 sequences from GenBank and generated 1384 new sequences for up to 15 genes, including six mitochondrial (12S, 16S, cox1, cytb, nd2, nd4) and nine nuclear (amel, bdnf, c-mos, jun, hoxa13, nt3, r35, rag1, rag2) loci, for a total of 6447 sequences for 15 genes (S3 Table). More than 640,000 nucleotide sequences for snakes are currently deposited in GenBank, representing an unparalleled resource for studies of the genetic diversity of the group. However, the quality and reliability of these data are a concern because of misidentification, mislabelling, and sequence contamination which seem to be the principal sources of error present in public databases [39]. In our search for sequences of Caenophidian snakes deposited in GenBank, we found 38 problematic sequences that were not included in the present analysis. All questionable sequences are reported in the supporting information (S6 Table) with succinct descriptions of the possible problems affecting each rejected sequence. New sequences generated in this study represented 21% of our whole matrix (S3 Table), with more than 50% and up to 80% of new sequences added to previously known sequences for genes amel, bdnf, nt3, jun, and hoxa13, while no sequences were generated for cox1, nd2, nd4, r35, and rag2 (S3 Table). Newly added sequences are illustrated with colored diamonds on each tip of terminals, representing the percentage of data generated in this study (white, 0%; light gray, between 1% and 50%; dark gray, between 50% and 99%; black, 100%). We did not sample all terminals for all genes, and the percentage of missing terminals varied from 16% for cytb to 93% for amel (S3 Table). The total number of species for families and subfamilies recognized in Uetz et al. [ 33 ] are summarized as supporting information (S5 Table). Tissue samples were obtained from museum collections. 7 / 82 According to Uetz et al.?s [ 33 ] generic and specific taxonomic listing, our complete taxon sampling of colubroids includes 78% and 42% of all recognized genera and species, respectively, totaling 344 genera and 1263 species (S4 Table). From this total of 1263 species, 20 species are recognized here but not listed by Uetz et al. [ 33 ] (S7 Table). According to these slightly revised numbers, summarized in S4 Table, we have the following generic and specific representations in percentage, respectively: Colubridae (77% and 37%), Dipsadidae (81% and 31%), Elapidae (85% and 54%), Homalopsidae (57% and 47%), Lamprophiidae (83% and 42%), Natricidae (68% and 39%), Pareidae (100% and 60%), Pseudoxenodontidae (100% and 40%), Viperidae (97% and 73%), Xenodermidae (67% and 28%). Outgroup sampling included representatives of the following families (number of terminals in parenthesis): Acrochordidae (3), Aniliidae (1), Boidae (2), Bolyeriidae (1), Calabariidae (1), Cylindrophiidae (1), Erycidae (1), Loxocemidae (1), Pythonidae (1), Uropeltidae (1) and Xenopeltidae (1). Trees were rooted with the typhlopid Indotyphlops braminus. DNA sequencing DNA was extracted from scales, shed skin, liver or muscle tissues using the phenol:chloroform method following specific protocols for each tissue [40,41]. PCRs were performed using standard protocols [41] for 11 genes, including four mitochondrial (12S, 16S, cox1, cytb) and seven nuclear (amel, bdnf, c-mos, jun, hoxa13, nt3, rag1). The sequences for each pair of primers and their respective references are provided as supporting information (S8 Table). PCRs were purified with shrimp alkaline phosphatase and exonuclease I (GE Healthcare, Piscataway, NJ). Sequences were generated in Brazil at the Laborato?rio de Biologia Geno?mica e Molecular, Pontif??cia Universidade Cato?lica do Rio Grande do Sul (Porto Alegre, Rio Grande do Sul) using the DYEnamic ET Dye Terminator Cycle Sequencing Kit in a MegaBACE 1000 automated sequencer (GE Healthcare); and in China at Laboratory for Conservation and Utilization of Bio-resources, Yunnan University (Kunming, Yunnan) using BigDye Terminator cycle sequencing kit in an ABI 3700 sequencer (Applied Biosystems, Foster City, CA). Both strands were sequenced for all fragments and sequences were edited and assembled using Geneious 5.5 (http://www.geneious.com) [42]. Phylogenetic analysis Sequences were aligned using MAFFT version 6 [43] applying the E-INS-i algorithm for rRNAs (12S and 16S) and the FFT-NS-i algorithm for protein coding sequences. The scoring matrix for nucleotide sequences was set to 200PAM/k = 2 and gap opening penalty was set to 1,53. Because sequences from GenBank present significant differences in size, we aligned them using a specific procedure that accounts for blocks of overlapping sequences to avoid alignment errors in both extremities of the aligned sequences. Extremities were then realigned using the same algorithm previously applied for each separate gene. Although our matrix retains high levels of missing data (average of 77.9%), sequences data per taxon range from 286 to 12659 bp with an average of 3299 bp. Similarly, highly incomplete taxa have been argued to be of minor concern in large-scale analyses that include many informative characters [ 44?45 ] and, as elsewhere [ 26 ], our most highly incomplete taxa were consistently placed in phylogenetic positions that are similar to previous works. We used PartitionFinder v1.1.1 [ 46 ] in order to select a partition scheme and evolutionary models based on AICc. We used the program RAxML version 7.2.8 [47] to perform a phylogenetic analysis employing Maximum Likelihood (ML) as the optimality criterion. We ran 1000 pseudoreplications of non-parametric bootstrap and we calculated FBP [ 48 ] using the rapid bootstrap algorithm implemented in RAxML (-f a). This also conducts a search for the ML 8 / 82 tree using each 5th bootstrap tree as a starting point for the rapid hill-climbing search (totaling 200 starting trees). Based on the 1000 trees derived from the pseudoreplications we calculated the TBE [ 49 ] using RAxML-NG [ 50 ]. We also calculated branch support using SHL [ 51 ] as implemented in RAxML (option?f E) for each branch of the tree. Comparing measures of clade support We compared SHL, FBP, and TBE for clades in our molecular tree and used them in combination to evaluate the robustness of specific clades [ 48,49,51 ]. We choose to use only the joint values of SHL and FBP to comment the results and base our discussion because they produced more conservative values than TBE (see results on comparisons of support metrics below for more details). We classified the robustness of each clade in seven categories based on the combined clade supports given by the SHL/FBP pair of support measures. These categories are graphically illustrated as supporting information (S2 Table) and summarized on the upper left corner of figures in the text, and are described as follows: 1) unambiguously supported, when both support methods recover values of 100%; 2) robustly supported, when clade support is not unambiguous, but both methods recover values 90%, or 80% in one method and 100% in the other; 3) strongly supported, when clade support does not reach percentages equal to previous categories 1 and 2, but both methods recover values 80%, or values 70% in one method and 90% in the other; 4) moderately supported, when clade support does not reach percentages equal to previous categories 1, 2, and 3, but both methods recover values 70%; 5) ambiguously supported, when clade support presents highly discrepant values, with < 70% in one method and 80% in the other method; 6) poorly supported, when clade support presents values < 70% in one method and between 70% and 80% in the other method; 7) unsupported, when clade support presents values < 70% for both methods (S2 Table). Although recognizing the subjectivity and arbitrariness of the described categories, we apply this approach in order to clearly state our reasoning and to facilitate the description of our general level of confidence in each clade retrieved by our phylogenetic analysis. Based on our seven categories, we suggest two main groups of combined support values: a first one with unquestionable or confident combined support values (categories 1, 2, 3, and 4), and a second one with contradictory or unsatisfactory support values (categories 5, 6, and 7). We highlight clades with contradictory (ambiguous) FBP/SHL support values because we consider them to be potentially erroneous and thus problematic when used in taxonomy or as presumptions for studies applying phylogenetic comparative methods (traits evolution), estimations of diversification rates (speciation/extinction rates) approaches of historical biogeography (discovery and event-based biogeography) and other methods requiring estimates of phylogeny. These unsupported clades should be treated with caution and either the uncertainty taken into account or their use eschewed altogether. Morphological comparisons We revised key morphological characters from the skull, hemipenis and vertebrae in representatives of most extant caenophidian families. Regarding the skull, we focused in two main anatomical complexes?the optic nerve foramen/fenestra and the naso-frontal joint?known to be phylogenetically informative in caenophidian snakes [ 30 ]. In that sense, we do not intend to provide here a thorough revision of colubroidean anatomy, and prefer to refer to Underwood [1], McDowell [ 8 ], Zaher [ 29 ], and Cundall and Irish [ 30 ] for a more complete review of the pertinent literature related to these morphological complexes. However, we provide figures of the relevant views of the skulls (S1 Appendix), vertebrae (S2 Appendix), and hemipenes 9 / 82 (S3 Appendix) from representatives of most caenophidian groups in order to illustrate the character states discussed herein. Specimens examined are listed in their respective supporting information files (S1?S3 Appendices). Institutional acronyms are as follow: AMNH, American Museum of Natural History, New York; BMNH, The Natural History Museum, London; FMNH, Field Museum of Natural History, Chicago; HUJR, Museum of Zoology, Hebrew University, Jerusalem; IBSP, Instituto Butantan, S?o Paulo; KU, Museum of Natural History, University of Kansas, Lawrence; LSUMZ, Louisiana State University, Baton Rouge; MNHN, Muse?um national d?Histoire naturelle de Paris; MZUSP, Museu de Zoologia, Universidade de S?o Paulo; ROM, Royal Ontario Museum, Toronto; UMMZ, Museum of Zoology, University of Michigan, Ann Arbor; USNM, National Museum of Natural History, Washington; ZMB, Museum fu?r Naturkunde, Berlin. Divergence time estimates We generated a time-calibrated tree for our complete molecular data set that provided a framework for the interpretation of paleontological, biogeographic, and cladogenetic patterns of caenophidian diversification. The large size of our molecular matrix (1278 terminals and 15 genes) precluded the use of commonly available parametric uncorrelated relaxed clock methods, as implemented in BEAST [ 52 ] and PAML [53]. Instead, we used an autocorrelated relaxed clock method based on a penalized likelihood implemented in the program treePL [54,55]. Divergence times were calculated in treePL by applying a smoothing parameter that defined the penalty for shifting the evolutionary rates among branches. This semiparametric method has been very effective for other data sets with large numbers of taxa [56,57]. To determine the smoothing parameter, we iterated 20 times a cross-validation procedure based on the RSRCV method (random subsample and replicate cross-validation) [55] with lambda values ranging from 0.01 to 100,000 (select lambda = 10) and the thorough option to ensure that the run iterates until convergence. We used the multicore option to distribute the cross-validation analyses on 64 processors of a Linux server. Since treePL only implements uniform prior distributions for node calibration points [54,55], the option of setting an open uniform distribution (by not defining a hard lower bound) can have two main undesirable effects: 1) estimating unrealistic older divergence times; and 2) providing a much larger space for parameter sorting, and thus decreasing the level of convergence of the estimation. In order to avoid these problems, we set maximum ages based on a phylogenetic approach, which takes into consideration the age of relative cladogenetic events. We made the assumption that the maximum age of a specific clade cannot be older than the minimum age of its more inclusive clade. For the purpose of our analysis, two lower bound dates were used to set uniform prior maximum ages: 93.9 Mya. (split between Alethinophidia and Scolecophidia) as the maximum age for our calibration of non-colubroidean nodes, and 54 Mya. (split between Colubriformes and Xenodermidae) as the maximum age for colubroidean nodes. Calibration points, fossils and constraint dates chosen for our divergence time analysis were as follow: 1. Alethinophidia stem clade?Haasiophis terrasanctus Tchernov, Rieppel, Zaher, Polcyn & Jacobs, 2000 was set as the Most Recent Common Ancestor (MRCA) of Anilius scytale and Indotyphlops braminus. The holotype corresponds to a complete, articulated specimen recovered from the Ein Yabrud quarries, near Ramallah, West Bank Palestinian Territories (Hebrew University of Jerusalem Paleontological Collections, HJU-PAL EJ 695). Hsiang et al. [58] combined molecular and morphological data in an unconstrained 10 / 82 analysis analysis and reported that Haasiophis terrasanctus was a stem-Alethinophidia instead of a crown-Alethinophidia and we adopted their conclusion. New interpretations of several characters have supported this view [58]. Because Hsiang et al.?s [58] analysis incorporates the known phylogenetic uncertainties related to the position of this fossil, we prefer to use Haasiophis as a stem-alethinophidian instead of a crown-taxa as traditionally used. The unclear position of the Ein Yabrud quarries within the Cenomanian may fall either in the Early Cenomanian Bet-Meir Formation [59] or the Late Cenomanian Amminadav Formation [60]. We followed Head [61] in using an age range that reflects the uncertainty of the position of Ein Yabrud, by assuming the minimum age set for the Cenomanian [62]. Thus this root-node was set as a hard bound [54], and constrained to 93.9 Mya [61]. 2. Boinae stem clade?Titanoboa cerrejonensis Head, Bloch, Hastings, Bourque, Cadena, Herrera, Polly & Jaramillo, 2009 was set as the MRCA of Boa constrictor and Eryx colubrinus. The holotype corresponds to one single precloacal vertebra (UF/IGM 1), and referred material includes 185 additional precloacal vertebrae and associated ribs representing 28 individuals from the Cerrejo?n Coal Mine, Rancheria River Valley, Guajira Peninsula, Colombia [63]. Morphology of the paracotylar foramina and a convex anterior zygosphene margin suggested that Titanoboa cerrejonensis belongs to the Boinae [61,63,64]. Head?s [61] preliminary analysis of undescribed cranial elements placed it in the stem lineage of the boine radiation [65]. Titanoboa cerrejonensis was recovered from sediments located within the Palynological zone Cu-02 of the Cerrejo?n Formation, dated as Middle to Late Paleocene [66,67]. Thus, the minimum age was constrained to 58 Mya and maximum age constrained to 93.9 Mya. 3. Colubriformes stem clade?Procerophis sahnii Rage, Folie, Rana, Singh, Rose & Smith, 2008 was set as the MRCA of Asthenodipsas vertebralis and Achalinus rufescens. The holotype consists of one posterior precloacal vertebra ("Rana Collection" from Vastan, VAS 1014), and referred material includes five precloacal vertebrae and two caudal vertebrae from the Vastan Lignite mine, Gujarat, India [68]. The lightly built and elongate shape of the precloacal vertebrae, presence of tapering prezygapophyseal processes, and blade-like uniformly thin neural spine that reaches the roof of the zygosphene refer P. sahnii to the clade Colubriformes. The combination of these characteristics excludes Procerophis from an association with the families Acrochordidae, Russellophiidae, Anomalophiidae, and Xenodermidae (S2 Appendix). The differentiated para- and diapophysial articular facets further distinguishes Procerophis from russellophiids and anomalophiids. The presence of a plesiomorphic prezygapophyseal morphology, with articular facets predominantly anteriorly angled, supports a basal position within Colubroides [ 68,69 ]. The rich squamate fauna from the Vastan Mine of the Cambay Formation was recovered from thin continental lenses of dark claystone and underlying marine shell beds, indicative of a near-shore environment deposited about 1 m above one of the two major Lignite layers (Lignite 2) present in the mine [ 70 ]. The squamate layer is situated approximately 14 m below the occurrence of the age-diagnostic foraminiferan Nummulites burdigalensis burdigalensis [ 71,72 ], indicative of shallow benthic zone SBZ 10 of Middle Ypresian age, which defines a minimum age of ~53 Mya for the deposit [ 73?75 ]. However, we here follow [70] in constraining the age of the vertebrate bearing bed of Vastan mine to an early-middle Ypresian (~54 Mya). The occurrence in the section of dinoflagellate cysts of early Ypresian age (~54?55 Mya) [ 76 ] and Strontium isotope age estimates for the deposits based on 87Sr/86Sr values clustering at an age of 54 Mya [77] support this slightly older age. The minimum age was constrained to 54 Mya and maximum age constrained to 93.9 Mya. 11 / 82 4. Viperidae stem clade?Vipera cf. V. antiqua [78] was set as the MRCA of Azemiops feae and Causus lichtensteinii. A moderately well-preserved cervical vertebra (Staatliches Museum fu?r Naturkunde, Stuttgart, SMNS uncatalogued) from Weisenau, Germany. The presence of a long and straight, slightly posteroventrally directed hypapophysis and a large condyle with a ventral portion lying on the posterior margin of the hypapophysis refers unambiguously the cervical vertebra from Weisenau to the family Viperidae. However, its assignment to the "Vipera aspis complex" by Szyndlar and Rage [79] is questionable since characters that are known to be diagnostic of the "Vipera aspis complex", such as an elongate centrum and short neural spines, are not marked in the cervical region. Additionally, an elongate centrum and short neural spines have been reported in distantly related genus Causus [80] and more closely related Daboia mauritanica [81]. Furthermore, the traditional subdivision of the "Vipera aspis complex", as originally proposed by Saint Girons [82] and detailed by Nilson and Andre?n [83] and Herrmann and Joger [84], corresponds to a paraphyletic arrangement of species as evidenced in our analysis (S1 Fig). The vertebra from Weisenau, as well as those reported from Saint-Gerand-le-Puy [79], Hessler [85], and Amo?neburg [86], are all from the earliest Miocene of Europe (European Land Mammal Age Neogene units MN 1 and MN 2), with the former likely being the earliest record for the family (MN 1). Although the precise age of the Saint-Gerand-le-Puy and Hessler viperids have been ambiguously associated to deposits that may come from either MN1 or MN 2 [85?87], Weisenau vipers are still associated with MN 1 deposits [86]. Thus, the minimum age was constrained to 22.1 Mya (MN 1) based on the Weisenau viperid vertebra. Maximum age was constrained to 93.9 Mya. 5. Crotalinae stem clade?Crotalinae gen. & sp. indet. A [88] was set as the MRCA of Azemiops feae and Tropidolaemus wagleri. A left maxilla with an almost complete tooth preserved in position (Department of Paleozoology, Institute of Zoology, Kiev, Ukraine, IZAN 3748) recovered from karstic fillings within a limestone quarry near the village of Gritsev, Shepetovski district, Ukraine. The maxilla was assigned unambiguously to the Crotalinae due to the deep depression on its posterolateral surface for the accomodation of the thermoreceptive (pit) organ. Holman [89] and Holman and Tanimoto [90] reported possible crown Crotalines from the lower Miocene of the U.S.A. and Japan, respectively. However, despite the fact that there are no known records of viperines in Japan or the New World, the identity of these older records as either crown or stem crotalines cannot be unambiguously determined based on vertebral morphology alone. Thus, we retained the maxilla described by Ivanov [88] as the only unambiguous crotaline record. Ivanov [88] reported that the stratigraphic age of the site from Gritsev corresponds to the middle Sarmatian MN 9a of Western Europe, with a minimum age of 10.4 Mya [ 91,92 ]. However, Vangengeim and Tesakov [ 93 ] argued that Gritsev was more accurately placed within the upper part of the middle Sarmatian, which lies in a zone of reversed polarity that is correlated to chron C5r. Therefore, we used the minimum age of 11.2 Mya estimated for the boundary between the upper and middle Sarmatian and correlated with subchron C5r.1n. The maximum age was constrained to 54 Mya. 6. Elapidae stem clade?Elapid Morphotype A [ 94 ] was set as the MRCA of Naja naja and Buhoma depressiceps. One mostly complete posterior trunk vertebra (Tanzanian Antiquities Unit, RRBP 04320) from locality TZ-01, Rukwa Rift Basin, southwestern Tanzania. This specimen was referred to the family Elapidae due to its low and robust hypapophysis and absence of a postzygapophyseal foramen [ 94 ]. It also shares a hypapophysis with a flattened and laterally expanded ventral edge with some members of the genus Naja [ 2 ]. According to McCartney et al. [ 94 ], the snake-bearing sites come from fluvial facies that belong to the 12 / 82 Songwe Member of the Nsungwe Formation, which is temporally constrained to ~ 24.9 Mya by mammalian biostratigraphy [ 95?99 ], detrital zircon geochronology, and a radiometrically dated volcanic ashes [100?102]. Thus, we constrained the minimum age to 24.9 Mya. Maximum age is constrained to 54 Mya. 13 / 82 N. merkuriensis, N. sansaniensis) [117,118]. However, all of these taxa consisted of either vertebral material or a few inconclusive cranial elements (e.g., compound bone, and quadrate), and none retain well-preserved parabasisphenoids. Therefore, Natrix aff. longivertebrata is here considered an unquestionable natricid record and is placed within colubroids, as a crown Natricidae. The minimum age was constrained to 13.8 Mya and maximum age to 54 Mya. Results We compare our molecular data and phylogenetic results with three recently published largescale molecular studies: Pyron et al. [ 26 ], Zheng and Wiens [ 27 ], and Figueroa et al. [ 28 ]). Pyron et al.?s [ 26 ] and Zheng and Wiens? [ 27 ] included representatives of all recognised squamatan families, whereas Figueroa et al. [ 28 ] focused on snake lineages. Caenophidian coverage in Pyron et al. [ 26 ] and Zheng and Wiens [ 27 ] were identical (1062 species) and included sequences from up to 12 and up to 52 genes, respectively. Figueroa et al. [ 28 ] combined up to 10 genes for 1358 species of caenophidian snakes (excluding multiple individuals, unidentified, and misidentified species). Our study combines sequences from up to 15 genes for 1263 species (see S3 Table for number of genes and accession numbers; see also S4 and S5 Tables for more details on taxon sampling). Higher-level relationships in these four large-scale studies are illustrated in Figs 1?3 and discussed below. The comparisons among support metrics are discussed below and illustrated in Figs 4 and 5. We also describe and illustrate separately the tree topology we obtained for each well-supported colubroidean family (Figs 6?21). The full tree (including outgroups) is provided as supporting information (S1 and S2 Figs). FBP and SHL, respectively, are provided in parenthesis for each recovered clade discussed below and in Figs 6?21. When applicable, the percentage of valid species sampled for a given genus is also shown in parentheses after the name of the genus (see S4 and S5 Tables for a summary and a list of sampled species per genus). 14 / 82 Fig 4. Scatterplots comparing support metrics for internal branches in the Maximum likelihood species-level phylogeny of Colubroides. A) TBE and FBP B) SHL and FBP C) SHL and TBE, D) Histogram showing the proportion of each category of joint support in each comparison of support metrics, E) Categories of joint support. Comparison of support metrics Support scores are strongly but imperfectly correlated. Pairwise scatterplots (Fig 4) show that TBE and SHL scores tend to be higher (less conservative) than FBPs (see also S9 Table). Thus, combining FBP and SHL produces almost the same proportion of supported clades as does combining FBP and TBE, whereas combining TBE and SHL increases the number of seemingly well-supported clades. This suggests that SHL and TBE tend to inflate and/or FBP tends to underestimate the support for clades. Categories of combined support values can help express such discrepancies by classifying all ambiguous supports as weakly supported clades (gray points and bars in Fig 4). The tendency for SHL and TBE values to be higher than FBPs affects all clade ages (Fig 5; S3 Fig). FBP tends to have weaker support values especially for deeper nodes in our ML tree (S1 and S2 Figs). 15 / 82 Fig 5. Distribution of branch support scores for each node age based on the Maximum likelihood species-level phylogeny of Colubroides. A) FBP distribution, B) SHL distribution, C) TBE distribution. Red dots represent values greater than 70%; gray dots indicate values smaller than 70%. 16 / 82 Fig 6. Maximum likelihood species-level phylogeny of Colubroides. Families Xenodermidae, Pareidae, subfamily Viperinae. Skeleton of the complete tree is displayed on the left, with the area of the tree corresponding to the present figure highlighted in black. Colored squares on each node represent bootstrap and SHL values following the categories of combined clade support described in S2 Table and summarized on the upper left corner of the figure. Diamonds on each tip represent the percentage of data generated in this study for each terminal: white, 0%; light grey, between 1% and 50%; dark grey, between 50% and 99%; black, 100%. 17 / 82 Fig 7. Maximum likelihood species-level phylogeny of Colubroides (continued). Family Viperidae, subfamilies Viperinae, Azemiopinae, Crotalinae. 18 / 82 Fig 8. Maximum likelihood species-level phylogeny of Colubroides (continued). Family Viperidae, subfamily Crotalinae. https://doi.org/10.1371/journal.pone.0216148.g008 PLOS ONE | https://doi.org/10.1371/journal.pone.0216148 May 10, 2019 19 / 82 Fig 9. Maximum likelihood species-level phylogeny of Colubroides (continued). Family Viperidae, subfamily Crotalinae. https://doi.org/10.1371/journal.pone.0216148.g009 PLOS ONE | https://doi.org/10.1371/journal.pone.0216148 May 10, 2019 20 / 82 position is not statistically supported in our molecular tree, and it might be equally possible for this species to cluster with Calamariidae since none of the branches that separate them received significant combined SHL/BH support values (Fig 19). Therefore, in face of the hemipenial and osteological similarities shared between Oreocalamus, Calamaria, and Macrocalamus, we assign it to the Calamariidae instead of the Colubridae. Colubroelaps?The monotypic genus Colubroelaps was described by Orlov et al. [163] to accomodate a small fossorial snake from southern Vietnam. They provisionally included the new genus in the family Colubridae (their subfamily Colubrinae) based mainly on the absence of hypapophyses on the posterior trunk vertebrae and of a diastema and sulcate teeth on the posterior end of the maxillae. However, the skull morphology of the type specimen shows that C. nguyenvansangi has hinged teeth like sibynophiids (Fig W in S1 Appendix). Among colubroidean snakes, Liophidium and Iguanognathus also retain hinged teeth [ 25 ], but differently from the latter two genera, Colubroelaps shares with sibynophiids the presence of a distally broadened, plate-like maxillary process of the palatine, absence of a choanal process of the palatine, a long tubular dorsally-curved compound bone, reduced mandibular fossa, vestigial splenial and angular bones, and a posterior dentigerous process of the dentary separated from the compound bone and forming a projected free ending process that diverges from the main mandibular axis (Figs V and W in S1 Appendix). Also, like Sibynophis and Scaphiodontophis, the maxilla of Colubroelaps projects freely posteriorly to the maxillary-ectopterygoid contact, but without forming an elongated ?dentigerous process? [ 25 ]. The combination of these derived characters shared by Sibynophis, Scaphiodontophis, and Colubroelaps supports the inclusion of the latter genus in the family Sibynophiidae. The absence of hypapophyses on the posterior trunk vertebrae of C. nguyenvansangi seems to contradict the present allocation because sibynophiids retain well-developed hypapophyses throuhought the trunk vertebrae (Fig G in S2 Appendix). However, we suspect that posterior hypapophyses are reduced due to the fossorial habits of that species. Iguanognathus?The genus Iguanognathus was tentatively allocated in the families Colubridae and Natricidae in the past, despite the lack of any compelling evidence supporting either hypotheses. Like sibynophiids, Iguanognathus has hinged teeth [164], which would suggest a possible close relationship with that family. However, apart from the presence of hinged teeth, Iguanognathus does not share any of the other cranial specializations typical of sibynophiids [ 25 ], as discussed above. Instead, Iguanognathus retains an aligned posterior dentigerous process of the dentary, well developed, functional splenial and angular bones, a choanal process, and a posteriorly curved tapering maxillary process of the palatine. Additionally, unlike sibynophiids, Iguanognathus lacks hypapophyses on the posterior precloacal vertebrae [165] which also rules out its belonging in the Natricidae. For these reasons, we prefer to consider this genus a Colubridae incertae sedis. A time-calibrated tree and the fossil record of Colubroides Previous age estimates of Colubroides. Divergence time estimates have been increasingly discussed in molecular studies of snake evolution in recent years [ 26,61,69,58,132,133,166 ]. Although they have been used to detail the tempo and mode of evolution of the group, these studies have sometimes inferred substantially different dates for major events in colubroid diversification. As Fig 23 shows, nine recent studies provide disparate dates for the origin of most higher-level clades of colubroideans [22,58,124, 132?135,166]. Our divergence time estimates are mostly concordant with those of Burbrink and Pyron [132] but are significantly younger than the dates estimated by some other studies [ 26,58 ]. Although there are differences in which fossils were used for calibrations, and in the numbers of genes and taxa, the different 61 / 82 results in Fig 23 mostly reflect expected differences between analyses based on autocorrelated and uncorrelated molecular clocks [167]. Studies in which time estimations were generated by penalized likelihood algorithms (e.g. treePL) tend to keep the same general pattern and same temporal cladogenic order. Comparing only the studies using autocorrelated methods [ 26,22,132 ] and our own (Fig 23), we observe that a difference among time estimations for one specific clade implies differences in the entire cladogenic process, that can be younger or older as a whole, but can never be younger for some clades and at the same time older for others, or vice-versa. In contrast, time estimations generated by uncorrelated methods [58,124,133,134] and a Bayesian autocorrelation method [135] tend to be more variable with respect to the general pattern, presenting cladogenic events that are ordered different among each study. As an example, a previous analysis of the family Viperidae using an uncorrelated method [124] in BEAST resulted in much older cladogenic events for the family as a whole (Fig 23). These differences likely result from the different methods of divergence time estimation used by Alencar et al. [124] and the present study. Such disparate results suggest that inferred dates of divergence should be treated with caution, and that the available fossil evidence is paramount to an accurate description of the evolutionary trends of a group. Therefore, we integrate our estimated divergence dates with the fossil record in an attempt to reach more balanced conclusions regarding the evolutionary events underlying the origin and diversification of extant colubroidean families. Despite their differences, some general trends emerge from the eight studies illustrated in Fig 23. Nine of the ten studies estimated an early divergence time for the ancestor of Colubroideans (i.e., the split between Colubroidea and Elapoidea) which, together, span an interval of approximately 35 My, from the upper Cretaceous (Turonian) to the upper Paleocene (Thanetian). Six of these studies place the origin of the group within the Cretaceous while three retrieve a Paleocene origin. The former hypothesis of a Cretaceous origin of the group is concordant with the presence of alleged colubroidean vertebral remains in the Cenomanian of the Wadi Milk Formation of Sudan [168]. However, the more complete material described by Rage and Werner [168] belongs to the enigmatic Caenophidian family Russellophiidae, a group known only from vertebral remains and only tentatively assigned to the clade Colubroides. The other colubroidean vertebrae were considered of indeterminate Colubroidean affinities due to their fragmentary condition [168], lacking preserved parts with unambiguous derived colubroidean traits and rendering their assignment to this group questionable [169]. Although a late Cretaceous origin of the group seems likely, more definitive evidence of Cretaceous colubroidean records is lacking. Colubroidean early divergence. The colubroidean fossil record is mostly composed of disarticulated vertebrae that are difficult to assign to any extant family because vertebral characters alone are of limited value when it comes to diagnosing most colubroidean clades [ 31,113 ]. Despite this limitation, colubroidean precloacal vertebrae were recognized until recently by the following suite of derived characters, known to occur in combination only in colubroidean snakes [ 31,68,109,169 ]: an elongate centrum, well-developed prezygapophyseal accessory processes, a blade-like neural spine that extends anteriorly onto the zygosphene and remains uniformly thin anteroposteriorly (as opposed to an expanded posterior margin of the neural spine), well-developed subcotylar tubercles (or cotylar ventrolateral processes), distinct dia- and parapophyseal articular facets of the synapophysis, prominent hypapophyses on middle and posterior trunk vertebrae, and paracotylar foramina. Among these characters, the uniformly thin blade-like neural spine that extends onto the roof of the zygosphene appears to be invariably present in all colubroideans, and unique to the group. However, our observations reveal that extant xenodermids lack an uniformly blade-like neural spine that reaches the roof ot the zygosphene (Fig 25; Fig A in S2 Appendix) [ 30,31 ]. 62 / 82 Although Achalinus, Xenodermus, and Fimbrios tend to retain a blade-like posterior margin, the neural spine never invades the roof of the zygosphene anteriorly [ 30,31 ] (Fig 25; Fig A in S2 Appendix). Therefore, we consider this character to represent a putative synapomorphy of the clade Colubriformes, with important implications in the definition of the minimum age used as calibration point for the base of our estimated colubroidean divergence time tree. Among extinct putative caenophidian families Russellophiidae, Nigerophiidae, and Anomalophiidae, only the latter seems to retain a similar neural spine morphology [ 31 ], and might well represent an early colubriform lineage. However, the enigmatic nature of these three families, known only from sparce vertebral material, precludes any unambiguous allocation to the colubroidean radiation. Therefore, the earliest records of allegedly uncontested colubroidean vertebrae from the Lower to Upper Eocene [68,109,169?175] are more accurately assigned to the clade Colubriformes instead. Our time calibrated tree places the divergence of stem-colubroideans at ~ 56 Mya, near the Paleocene/Eocene boundary, while stem-colubriforms diverged at ~ 53 Mya within the Ypresian (Fig 22). Early fossil records of definitive colubriforms are concordant with this date, with the oldest unequivocal record being of Procerophis sahnii [68] from the early Ypresian of India, with an age of 54 Mya. (Fig 22) [ 68,71,72 ]. The other known Eocene colubriform records are all from the middle/upper Eocene: an unnamed colubriform from the middle Eocene of Namibia (41.2 Mya) [174]; an unnamed colubriform from the middle Eocene of Myanmar (37.2 Mya) [169]; Renenutet enmerwer from the middle Eocene of Egypt (37 Mya) [172]; a vertebra referred to Nebraskophis from the upper Eocene of Hardie Mine, USA (34.2 Mya) [171]; and an unnamed colubriform from the upper Eocene of Thailand (34 Mya) [175]. Because the vertebrae of Vectophis wardi and Headonophis harrisoni from the upper Eocene of the Isle of Wight, England (33.9 Mya) [170,173] retain a robust and posteriorly expanded neural spine that does not invade the zygosphenal roof anteriorly, we treated them as caenophidians of uncertain affinities instead of belonging to the clades Colubroides or Colubriformes. Molecular evidence supports an early Paleogene divergence of colubroideans in Asia [132,134], but they may have been present already in Africa in the early Upper Cretaceous [168]. The presence of a definitive colubriform snake in the Lower Eocene of Namibia [174], and the recent finding of Renenutet enmerwer in the Upper Eocene of Egypt [172] along with an already well established colubroidean fauna in the Lower Oligocene of Tanzania [ 94 ], indicates that the group had already diversified in Africa by the Eocene. Additionally, the presence of diversilly significant number of colubriform records in India during the Eocene, including the oldest undisputed colubriform snake, along with the African records discussed above, suggest that colubroideans may have diversified much earlier in Gondwana prior to its dispersal throughout Laurasia [172]. In that context, the controversial presence of colubroidean snakes in the Cenomanian of Sudan [ 69 ], which extends the divergence timing of the group to the early Upper Cretaceous, seems to become a plausible hypothesis [168]. However, additional findings are needed to fill the implied ghost lineage of approximately 40 million years, from the Upper Cretaceous sediments of the Wadi Milk Formation of Sudan to the Lower Eocene undisputed colubriform record of India. The Eocene-Oligocene transition and the diversification of present-day Colubroid and Elapoid lineages The early Oligocene was marked by a much cooler and more temperate global climate than the warm "greenhouse" conditions that characterized most of the Cretaceous and early Cenozoic [176,177]. The impoverished global diversity in the Oligocene that resulted from the Eocene63 / 82 Oligocene extinctions is also observed in the fossil record of snakes around the world [ 32 ]. Our time calibrated tree for colubroideans illustrates this trend, with a relatively low number of cladogenetic events dated prior to the Oligocene-Miocene boundary (Fig 22; S4 Fig). However, these Oligocene cladogenetic events were key for the establishment of the present-day colubroidean snake fauna. While an early divergence of basal colubroidean lineages is estimated to have occurred within the Eocene, all extant families of Colubroidea and Elapoidea that compose most of the present-day endoglyptodont fauna are estimated to have originated rapidly within the early Oligocene interval, between ~ 33 and 28 Mya. (Figs 22 and 23; S4 Fig). Diversification dates retrieved here are consistent with the major terrestrial faunal turnover recorded around the world and associated with the overall climate shift at the Eocene-Oligocene boundary. This trend is consistent with the sudden appearance in the European fossil record of the derived colubrid vertebral morphotype with an elongated centrum, long prezygapophyseal accessory processes, distinct epizygapophyseal spines, and a uniformly narrow haemal keel (lacking hypapophyses), as illustrated by Coluber cadurci [ 31, 69,110,109 ]. The subsequent, mainly Miocene diversification of extant Colubroidean families, is also highly consistent with the known Neogene colubroidean fossil record [32,178,179]. Among ?basal? colubroidean lineages estimated to have diverged within the Eocene (Fig 22), the Xenodermidae, Pareidae, and Homalopsidae still lack a fossil record, while the first unequivocal viperid record is only early Miocene of age (MN1) [79], contrasting significantly with our estimated timescale. Among Paleogene fossil caenophidian snakes, Thaumastophis missiaeni approaches the xenodermid vertebral morphology in having vertically oriented blade-like prezygapophyseal accessory processes and a lightly built and elongate vertebral morphology [68]. However, the combination of these two characters, along with the absence of well-developed hypapophyses (present in xenodermids), and the presence of parazygantral foramina (shared with acrochordids) and a blade-like neural spine invading the zygosphenal tectum (shared with colubriformes) precludes its assignment to any of the three colubroidean families cited above, or to the acrochordids [68]. The lack of a well-established fossil record for the Xenodermidae, Pareidae, and Homalopsidae during that interval of early colubroidean evolution hampers any attempt to determine in more details the pattern of early divergence of the group. Notwithstanding, according to our divergence estimates, it can be hypothesized that appearance of grooved venomous teeth and the consequent diversification of higher endoglyptodont lineages occurred within the Eocene, prior to the large-scale faunal turnover that characterizes the Eocene-Oligocene transition [176,177,180?183]. Indeed, our time calibrated tree indicates that stem-Xenodermidae diverged at ~ 52.6 Mya while stem-Pareidae and stem-Endoglyptodonta diverged at ~ 45 Mya. Stem-Viperidae and stem-Homalopsidae also diverged within the Eocene, at ~ 42.5 Mya and 38 Mya, respectively. Viperidae fossil record and divergence time estimates. Although viperids are abundant in the fossil record, most are confined to the Neogene, consist of isolated vertebrae, and are assigned to extant taxa [79,89]. The oldest record of a viperid known so far is Provipera boettgeri, described by Kinkelin [184] based on an isolated fang from the early Miocene of Germany (MN1; ~ 21 to 23 Mya) (see Rage [ 31 ] for the validity of the name). Viperids are also recorded in the early Miocene of Southern Asia (equivalent to MN 3) [185] and North America (late Arikareean) [186], early or middle Miocene of Central Asia [187], and middle Miocene of northern Africa (equivalent to MN 7+8) [188]. The first unquestionable crotaline was reported by Ivanov [88] from the middle Miocene of Gritsev in Ukraine (MN 9). Szyndlar and Rage [79,85] provided a detailed review of the known Neogene fossil record of the family. As shown by these authors, the fossil record of viperids does not help clarify the early divergence of the family, since most fossils are associated with extant taxa from derived lineages [79,85], as shown in our own phylogenetic tree (Figs 6?9). 64 / 82 Our time calibrated tree suggests that the origin of crown-Viperidae occurred in the early Oligocene, at ~ 30.7 Mya, while the basal split between viperine and causine subfamilies on the one hand, and crotaline and azemiopine subfamilies, on the other hand, occurred within the late Oligocene, at approximately 26 and 25 Mya, respectively (Fig 22). As such, although expected, no fossil viperids have been recorded yet in the Paleogene, resulting in an interval of approximately 20 Mya between their hypothesized early divergence in the Eocene and the first known, early Miocene, unequivocal fossil of the family [79,85]. Additionally, since the sistergroup relationship between Asiatic/American crotalines and African/Eurasiatic viperines is mainly symmetric, no conclusion can be reached on the geographic area of origin of the family (apart from excluding the New World). Elapoid fossil record and divergence time estimates. Within Elapoidea, only the Elapidae and, possibly, Pseudoxyrhophiidae have a fossil record [ 31,32,94 ]. The oldest known record of an unequivocal elapid comes from the late Oligocene of the Nsungwe Formation, Tanzania, which bears sediments of ~ 25 Mya [ 94 ]. McCartney et al. [ 94 ] also report a distinct caudal vertebra from the same locality in Tanzania that bears the unusual feature of a single hemal keel instead of paired hemapophyses. According to the authors, the extant genus Duberria also exhibits a similar condition, being the only known pseudoxyrhophiid snake so far that lacks hemapophyses but retains a well-developed hemal keel. Although Duberria?s caudal morphology ressembles the caudal vertebra described by McCartney et al. [ 94 ] as "Colubroid Morphotype C", the authors rightly refrain from assigning the latter to Pseudoxyrhophiidae. The oldest record of an Australian elapid consists of a vertebra attributed to a hydrophiine found in sediments of Riversleigh dated from the upper Oligocene or lower Miocene (24?23 Mya) [103]. According to Scanlon et al. [103] the vertebra is morphologically very similar to the extant genus Laticauda. However, these authors refrain in allocating it to the latter genus given the limited information afforded by one isolated vertebra. The first record of elapids in Europe comes from the lower Miocene of France (MN 4; [ 32 ]). Elapids are further abundantly documented throughout the Miocene and the Pliocene of Europe [ 115 ], persisting in that continent until the upper Pliocene when they became extinct [ 189?191 ]. According to Szyndlar and Rage [191], most European fossil elapids are assignable to extant Naja. This genus is also recorded in the middle Miocene of northern Africa (13.8 Mya; MN70 [ 188,191 ]). The relatively abundant skull material found associated with elapid vertebrae in the Neogene of Europe tend to corroborate the view that most of these large Neogene elapids were either closely related to or nested within Naja [ 115,190,192 ]. Isolated posterior trunk vertebrae from the middle Miocene of North America (upper Barstovian) and Europe (Astaracian, MN7) were assigned to extant Micrurus [ 193 ], as Micrurus sp. and Micrurus gallicus, respectively. Referral to the family Elapidae is based on having a low and recurved hypapophysis, a low anteroposteriorly elongated neural spine, and a poorly vaulted neural arch. However, these characters may correlate well with fossorial habits [ 193 ], and, thus, their assignment to the Micrurus is questionable. No compelling evidence distinguishes the vertebral morphology of Micrurus from other Asian and Neotropical coral snake genera. According to Rage and Holman [ 193 ], the few vertebrae referred to Micrurus from the Miocene of Nebraska (USA) and la Grive (France) are comparable to extant Micrurus fulvius, and thus should be referred to this genus. Our observations of the vertebral morphology of South American Micrurus shows a very distinctive morphology from that of Micrurus fulvius, suggesting that the vertebral morphology of the genus is much more diverse than previously thought. A detailed description of the vertebral morphology of speciose New World Micrurus and its closely related North American and Asiatic genera Micruroides, Leptomicrurus, Calliophis, and Sinomicrurus is necessary to confidently support the assignment of these Miocene records to any known extant genus, and especially to Micrurus. 65 / 82 Sub-Saharan elapids from the end of the Paleogene raise doubts about the well-accepted hypothesis of an Asian origin for the group [ 2,22,103 ]. According to McCartney et al. [94], the presence of elapids in the Nsungwe formation indicates two possible scenarios: a rapid initial phase of dispersion of the family from Asia to Africa before the end of the Oligocene or, alternatively, an origin of the family in Africa rather than Asia. The elapids from Nsungwe and the hydrophiine from Riversleigh help reduce the gap between the fossil record and the most recent molecular estimates (Fig 23). The presence of a hydrophiine in the late Oligocene or early Miocene of Australia further supports the hypothesis of a dispersal and colonization of the Australian continent in the late Oligocene [ 194 ]. Our time calibrated tree suggests an early divergence of stem-elapids within the early Oligocene at ~ 30.5 Mya, with the main crown-Elapidae lineages diversifying during the late Oligocene at ~ 26.5 Mya (Fig 22; S4 Fig). Although estimates of the time of divergence of Elapidae seem to favour an early Oligocene origin (Fig 22), available molecular phylogenies (including this one) and the fossil record do not yet allow inference of the biogeographic origin of the group. While the discovery of elapids in the upper Oligocene of sub-Saharan Africa and the unambiguous position of the family within the African elapoid radiation favour an African origin, the basal-most positions of successive Asian coral-snake lineages in recent molecular phylogenies tends to favor the opposite hypothesis of an Asian origin of the group. The lack of support for deeper elapid relationship fails to provide a robust support for either hypothesis (Figs 10?12). Apart from the uncertainties involving the debate on an Asian or African origin, it has been commonly thought that the family dispersed to the West Paleartic, Australian (via Melanesia), and Neartic/Neotropical (via the Bering strait) regions independently [ 2,32,103,115,183,194 ]. Scanlon et al. [103] provided robust paleontological evidence supporting the hypothesis of an over-water dispersal to Australia of the ancestor of the hydrophiine radiation, close to the Oligocene?Miocene boundary. The dispersal into the West Paleartic in the lower Miocene of forms belonging to or closely related to the extant genus Naja is also well documented [ 2,32,115,190?192 ]. In contrast, the occurrence of extant genus Micrurus in the Miocene of France is questionable due to the absence of vertebral diagnostic features that distinguish members of this genus from the Asiatic coral snake radiation. Its record in the Miocene of North America also needs further corroboration since Holman [ 195 ] used only two extant species of Micrurus (M. fulvius and M. affinis) and Micruroides euryxanthus for comparison. Our time calibrated tree places the early divergence of stem-hydrophiines at ~ 23 Mya, at the Oligocene?Miocene boundary, and the ancestor of the New World radiation of coral snakes (stem-micrurines) diverged from the Old World coral snakes at ~ 21 Mya in the early Miocene. These results support the hypotheses of a late Oligocene over-sea dispersal and colonization of the Australian continent by the ancestor of hydrophiines [103] and early Miocene terrestrial colonization of North America by the ancestor of New World coral snakes. Colubroid fossil record and divergence time estimates. The fossil record of Colubroidea is much more extensive than that of elapoids but is mostly confined to the Neogene. Their vertebral morphology remains poorly known, and an accurate evaluation of the fossils assigned to this superfamily or to specific colubroid families remains far from being resolved. Pseudoxenodontids, calamariids, sibynophiids, and grayiids have not been recorded so far in the fossil record. In contrast, ?colubrid? and natricid fossils are abundant and have been recorded throughout the Neogene of North America and Europe [ 31,175,179 ]. Colubrids and dipsadids are also well-represented in the Neogene of North America [178]. Although Miocene and Pliocene records of colubroids are straightforward, Paleogene records are more elusive and most of them are of uncertain assignment. Here we follow Smith [109] in assigning the vertebrae from the upper Eocene of the Medicine Pole Hills of the 66 / 82 Chadron Formation in North Dakota to Colubroidea since they retain a ?racer-like? vertebral morphology consistent with those of the North American racer clade of colubrids [109]. This record represents the oldest known Colubroidea so far. The divergence between Colubroidea and Elapoidea in our time calibrated tree is estimated at ~ 36 Mya, lying close to the boundary between the Eocene and Oligocene (Fig 22) and in accordance with the appearance of colubroids in the late Eocene of North America (Chadronian NALMA) [109]. Similar to elapoid families in our time calibrated tree, ancestors of extant families of colubroids diverged during the early Oligocene, between ~ 30 to 33 Mya. These dates are also in agreement with the first emergence of the typical colubrid vertebral morphotype in the early Oligocene of France, documented by Coluber cadurci from the Phosphorites of Quercy (Mammal Paleogene Reference Unit MP 22). Colubrid precloacal vertebrae can be distinguished from all other colubroidean family by the presence of the following combination of derived features: an elongated centrum, long prezygapophyseal accessory processes, distinct epizygapophyseal spines, and an uniformly narrow haemal keel (lacking hypapophyses). Slightly younger records of putative natricids from the Phosphorites of Quercy are difficult to allocate due to their fragmentary condition and their overall similarities with the vertebral morphology of several extant elapoid families. Holman [178] and Szyndlar [179] detailed the Neogene colubroid records in North America and Europe, respectively. Conclusions The traditional meaning of the superfamily Colubroidea [ 18,26,26,28 ] no longer accommodates our growing knowledge of the phylogenetic affinities [ 18,19,23,25,26,27,28 ] and morphological disversity and disparities in the group [ 8,11,23,25,29,30 ]. Despite some pleas against any change on the traditional usage of the names ?Colubroidea? and ?Colubridae? [ 196 ], the new morphological evidence provided here reinforces the need for a series of taxonomic changes to accommodate new phylogenetic and morphological knowledge. Xenodermids, pareids, and xylophiids represent ancient caenophidian lineages that are phylogenetically and morphologically distinct from the endoglyptodont radiation. All three lineages lack the dental specializations that gave rise to an advanced venom delivery system characteristic of endoglyptodonts, thus breaking a universally accepted definition of colubroids as representing the truly venomous snake radiation. Xenodermids further lack some vertebral and cranial features that are commonly used to determine colubroid fossil remains and share with acrochordoids cranial and vertebral specializations that are virtually absent in the remaining caenophidian lineages. Our observations on caenophidian vertebral morphology, especially in xenodermids, were particularly useful in redefining the fossil record commonly used as calibration points in molecular phylogenies (e.g., the oldest colubroidean remains?Procerophis sahni?is here placed as a calibration point for the early divergence between colubriformes and xenodermids instead of the traditional divergence between xenodermids and acrochordids). By reviewing and reinterpreting the relevant fossil record, our treePL analysis was able to highlight previously unnoticed correlation between the early diversification of colubroid and elapoid major lineages and the Eocene-Oligocene transition. Our divergence dates are in general younger than most previous studies (Fig 23) and, although many authors would suspect that disparate dates result from distinct methodological approaches, we suggest that most of these differences are due primarily to the effect of highly heterogeneous usages of the fossil record rather than of distinct methodological procedures. More importantly, our results highlight the need for more detailed anatomical studies in combination with a more careful usage of the fossil record [ 197 ]. 67 / 82 Similarly, comparative morpho-functional or behavioural studies with caenophidians, that are dependent on previously published, molecular phylogenies as frameworks, should seek more accurate, combined evidence of branch support to evaluate their evolutionary scenarios. Our statistical exploration of three different support methods indicates that TBE and SHL are constantly higher and less conservative than FBP values. Based on the discrepancies among these methods, we reinforce the combined use of different support values to identify nodes that are not well supported or ambiguously supported. Such approach can help highlight weakly supported clades and/or the presence of rogue terminals in phylogenetic datasets. Our study revealed that a large number of colubroidean clades are still either poorly or ambiguously supported and should be treated with caution. Supporting information S1 Table. Distinct classification schemes discussed in this study. Number of Phylogenetic levels in each classification scheme are as follow: 1?3 = higher-levels, 4 = superfamily, 5 = family, 6 = subfamily; families are listed in bold. (XLSX) S2 Table. Categories of combined clade support. Graphic illustration for combined clade support values when comparing FBP, SHL, and TBE metrics, classified in seven categories, as follows: 1) Red, unambiguously supported; 2) Orange, robustly supported; 3) blue, strongly supported; 4) green, moderately supported; 5) dark grey, ambiguously supported; 6) dark grey, poorly supported; 7) light grey, unsupported (see text for discussion). (XLSX) S3 Table. List of accession numbers. List of all the terminal taxa and sequences by gene used in the present study, including best partitions, accession numbers of sequences retrieved from GenBank and new sequences produced for this study. (XLSX) S4 Table. Number and percentage of species and genera sequenced by family. Comparisons between the number of families, genera and species of Colubroides used in the present study and by Figueroa et al. [ 28 ] (2016). (DOCX) S5 Table. Numbers and percentages for genera and species. Number of species by genus of Colubroides sampled in this study; and a list of all species of Colubroides following Uetz et al. [ 33 ], indicating which species was sampled in our study (spreadsheet 2). (XLSX) S6 Table. List of sequences from GenBank considered questionable and/or problematical. List of accession numbers, with genes names, current identification in GenBank, and probable correct identification for questionable and/or problematic sequences of snakes available in GenBank. (DOC) S7 Table. List of taxa recognized in this study but not listed by Uetz and Hosek (2017). List with rationale for the species recognized in the present study, but not listed in Uetz et al. [ 33 ]. (DOCX) S8 Table. List of Primers used in this study. List with sequences for the pairs of primers used to amplify the gene fragments used in the present study. (XLSX) 68 / 82 S9 Table. Numbers of clades in each category of combined FBP, TBE, and SHL support values. Clade support values based on the combination of (A) FBP/SHL, (B) FBP/TBE, (C) and TBE/SHL, classified in seven categories, as follows: red, unambiguous support (both methods recover values of 100%); orange, robust support (both methods recover values 90%, or 80% in one method and 100% in the other); blue, strong support (both methods recover values 80%, or values 70% in one method and 90% in the other); green, moderate support (both methods recover values 70% but do not reach values equal to previous categories); dark grey, ambiguous support (highly discrepant values, with < 70% in one method and 80% in the other) or poor support (values < 70% in one method and between 70% and 80% in the other method); light grey, unsupported (values < 70% for both methods). (XLSX) S10 Table. Representative fossil snakes from the Cenozoic. List of fossil snakes from the Paleogene and Neogene with authorship, stratigraphic occurrence and locality. (XLSX) S1 Fig. Full RAxML tree. Maximum likelihood tree of Colubroides containing 1263 terminals. Color of the squares follow the categories of combined clade support as described in S2 Table. Numbers inside de squares on the nodes of the full tree represent the bootstrap and SHL values retrieved. Diamonds on each tip represent the percentage of the data for each terminal generated in this study: white, 0%; light gray, between 1% and 50%; dark gray, between 50% and 99%; black, 100%. Terminals in red represent additional samples in relation to Pyron et al. [ 26 ]. (PDF) S2 Fig. Full treePL tree. Data matrix and calibrated tree resulting from the treePL analysis of Colubroides, including the outgroups and containing 1278 terminals (1263 Colubroides and 15 outgroups). (TRE) S3 Fig. Full RAxML tree. Maximum likelihood species-level phylogeny of Colubroides including comparisons among values of FBP, SHL, and TBE support metrics. Numbers inside de squares on the nodes of the full tree represent the TBE values retrieved. (PDF) S4 Fig. treePL zoomed trees. Zoomed, large-scale calibrated tree resulting from the treePL analysis showing the pattern of cladogenic events through time. (PDF) S1 Appendix. Skulls. Skull morphology of representatives of colubroidean families illustrating the naso-frontal joint and optic foramen/fenestra. Figure A, Tropidophiidae: Tropidophis nigriventris (AMNH 81182); Acrochordidae: Acrochordus granulatus (ZMB 9444). Figure B, Xenodermidae: Achalinus spinalis (AMNH 34621), Fimbrios klossi (BMNH 1946.1.15.88). Figure C, Xenodermidae: Xenodermus javanicus (FMNH 158613); Xylophiidae: Xylophis perroteti (BMNH 1955.1.3.10). Figure D, Pareidae: Pareas moellendorffi (AMNH 27770), Apopeltura boa (BMNH 47.12.30). Figure E, Viperidae: Azemiops kharini (ZMB 69985), Bothrops neuwiedi (MZUSP 1476), Causus rhombeatus (FMNH 74241), Vipera ursinii (MZUSP 8230). Figure F, Homalopsidae: Bitia hydroides (FMNH 229568), Brachyorrhos albus (FMNH 142322), Enhydris chinensis (AMNH 33870), Fordonia leucobalia (AMNH 107179). Figure G, Homalopsidae: Homalopsis buccata (MNHN 1963.728); Psammophiidae: Malpolon monspessulanus (AMNH 140768), Mimophis mahfalensis (UMMZ 209653). Figure H, Psammophiidae: Psammophylax variabilis (AMNH 73213), Rhamphiophis oxyrhynchus (AMNH 16890), 69 / 82 Psammophis phillipsi (AMNH 67750). Figure I, Cyclocoridae: Cyclocorus lineatus (MNHN 1900.413), Oxyrhabdium modestus (FMNH 53386); Atractaspididae: Aparallactus modestus (AMNH 50545). Figure J, Atractaspididae: Atractaspis bibronii (AMNH 82073), Homoroselaps lacteus (LSUMZ 57229), Macrelaps microlepidotus (FMNH 205860), Polemon christyi (FMNH 219913). Figure K, Lamprophiidae: Bothrolycus ater (AMNH 11976), Chamaelycus fasciatus (BMNH 1909.4.29.3), Dipsadoboa weileri (AMNH 12472), Lamprophis olivaceus (AMNH 12003). Figure L, Lamprophiidae: Lycodonomorphus rufulus (AMNH 140284), Lycophidion capense (AMNH 63771), Gonionotophis capensis (AMNH 73208), Pseudoboodon lemniscatus (MNHN 1905.179). Figure M, Pseudoxyrhophiidae: Alluaudina bellyi (UMMZ 201605), Dromicodryas quadrilineatus (UMMZ 209290), Duberria lutrix (UMMZ 154361), Heteroliodon occipitalis (UMMZ 218178). Figure N, Pseudoxyrhophiidae: Ithycyphus miniatus (UMMZ 201615), Langaha madagascariensis (UMMZ 218193), Liophidium torquatum (UMMZ 209437), Pseudoxyrhopus tritaeniatus (UMMZ 203649). Figure O, Elapidae: Bungarus caeruleus (AMNH 87483); Calliophis intestinalis (BMNH 1880.9.10.15), Micrurus narduccii (MZUSP 8370), Naja naja (AMNH 86912). Figure P, Elapidae: Notechis scutatus (ZMB 7930), Toxicocalamus loriae (AMNH 95581); Pseudoxenodontidae: Pseudoxenodon stricticaudatus (AMNH 34674). Figure Q, Natricidae: Afronatrix anoscopa (MNHN 1986.1618), Aspidura trachyprocta (AMNH 120251), Atretium schistosum (AMNH 85509), Lycognathophis seychellensis (UMMZ 195836). Figure R, Natricidae: Natriciteres fuliginoides (MNHN 1987.1552), Natrix maura (AMNH 115697), Sinonatrix annularis (AMNH 115693), Xenochrophis cerogaster (AMNH 89276). Figure S, Dipsadidae: Apostolepis cf. nelsonjorgei (MZUSP 20636), Atractus maculatus (IB 40003), Conophis pulcher (AMNH 117934), Contia tenuis (UMMZ 133519?1). Figure T, Dipsadidae: Farancia abacura (KU 214419), Geophis hoffmanni (AMNH 113561), Helicops pastazae (AMNH 49143), Heterodon nasicus (MNHN 1993.1625). Figure U, Dipsadidae: Philodryas mattogrossensis (AMNH 141377), Sibon sartorii (LSUMZ 23243), Tachymenis peruviana (KU 135193), Urotheca multilineata (AMNH 98284). Figure V, Dipsadidae: Xenopholis scalaris (AMNH 60799); Sibynophiidae: Scaphiodontophis annulatus (MZUSP 5971), Sibynophis subpunctatus (AMNH 96073). Figure W, Sibynophiidae: Colubroelaps nguyenvansangi (ZISP/IEBR 25682). Figure X, Calamariidae: Calamaria gervaisi (AMNH 36744), Macrocalamus lateralis (LSUMZ 45407); Oreocalamus hanitschi (BMNH 1929.12.22.106). Figure Y, Grayiidae: Grayia smithii (AMNH 140428); Colubridae: Boiga dendrophila (AMNH 116014), Coluber constrictor (FMNH 135284). Figure Z, Colubridae: Dendrelaphis papuensis (AMNH 107175), Ptyas mucosus (AMNH 83993), Scaphiophis albopunctatus (AMNH 104101), Senticolis triaspis (AMNH 110625). Figure AA, Colubridae: Spilotes pullatus (IBSP 4955); Colubridae incertae sedis: Iguanognathus werneri (BMNH 1946.1.6.34). Figure AB, Elapoidea incertae sedis: Buhoma depressiceps (BMNH 1907.5.22.10), Micrelaps muelleri (HUJR 8009). Scale bar = 1 mm. (PDF) S2 Appendix. Vertebrae. Posterior trunk vertebral morphology of representatives of colubroidean families. Figure A, Acrochordidae: Acrochordus javanicus (USNM 297404), scale bar = 2 mm; Xenodermidae: Achalinus rufescens (BMNH 1946.1.12.37), scale bar = 1 mm; Fimbrios klossi (BMNH 1946.1.15.88), scale bar = 1 mm; Pareidae: Pareas sp. (MZUSP 12186), scale bar = 1 mm. Figure B, Viperidae: Causus difilippi (MZUSP 18668), scale bar = 5 mm; Vipera ursinii (MZUSP 8230), scale bar = 5 mm; Azemiops feae (ROM 36976), scale bar = 1mm; Bothrops jararaca (MZUSP 14425), scale bar = 2mm. Figure C, Homalopsidae: Cerberus rynchops (MZUSP 9569), scale bar = 2mm; Homalopsis buccata (MZUSP 11483), scale bar = 1mm. Psammophiidae: Psammophis lineolatus (MZUSP 8221), scale bar = 1mm; Mimophis mahfalensis (MZSUP 12188), scale bar = 2mm. Figure D, Pseudoxyrhophiidae: Madagascarophis 70 / 82 colubrinus (BMNH 89.8.28.23), scale bar = 2mm; Ditypophis vivax (BMNH_99.12.5.125), scale bar = 1mm; Lamprophiidae: Boaedon fuliginosus (MZUSP 8167), scale bar = 2mm; Crotaphopeltis hotamboeia (MZUSP 19602), scale bar = 1mm. Figure E, Atractaspididae: Atractaspis irregulares (MZUSP 10826), scale bar = 1mm; Homoroselaps lacteus (LSUMZ 57229), scale bar = 1mm. Elapidae: Sinomicrurus macclellandi (ROM 37113), scale bar = 1mm; Naja naja (UMMZ 181137), scale bar = 1mm. Figure F, Elapidae: Micrurus corallinus (MZUSP 13112), scale bar = 1mm; Cyclocoridae: Cyclocorus lineatus (BMNH 96.3.30.78), scale bar = 1mm; Natricidae: Natrix natrix (MZUSP 2514), scale bar = 2mm; Natriciteres olivacea (MZUSP 2083), scale bar = 1mm. Figure G, Sibynophiidae: Scaphiodontophis annulatus (MZUSP 5971), scale bar = 2mm; Grayiidae: Grayia smithii (MZUSP 8130), scale bar = 1mm; Grayia tholloni (MZUSP 8135), scale bar = 2mm; Calamariidae: Oreocalamus hanitschi (BMNH 1929.12.22.106), scale bar = 1mm. Figure H, Colubridae: Chironius bicarinatus (MZUSP 13860), scale bar = 2mm; Spilotes pullatus (MZUSP 13845), scale bar = 2mm; Oxybelis aeneus (MZUSP 13028), scale bar = 2mm; Mastigodryas boddaerti (MZUSP 13052), scale bar = 2mm. Figure I, Colubridae: Simophis rhinostoma (MZUSP 13858), scale bar = 2mm; Dipsadidae: Heterodon platirhinos (MZUSP 2991), scale bar = 2mm; Farancia abacura (MZUSP 2953), scale bar = 2mm; Carphophis amoenus (MZUSP 8183), scale bar = 1mm. Figure J, Dipsadidae: Synophis lasallei (MZUSP 7713), scale bar = 1mm; Nothopsis rugosus (MZUSP 7490), scale bar = 1mm; Dipsas indica (IBSP 40137), scale bar = 1mm; Atractus serranus (MZUSP 17937), scale bar = 1mm. Figure K, Dipsadidae: Boiruna maculata (MZUSP 703), scale bar = 2mm; Helicops angulatus (MZUSP 14234), scale bar = 2mm; Philodryas nattereri (MZUSP 13039), scale bar = 2mm; Oxyrhopus clathratus (MZUSP 14010), scale bar = 2mm. (PDF) S3 Appendix. Hemipenes. Hemipenial morphology of representatives of colubroidean families. Figure A, Acrochordidae: Acrochordus javanicus (LSUMZ 34406) completely everted and filled, scale bar = 5 mm; Xenodermidae: Xenodermus javanicus (FMNH 138678) partially everted, partially filled, and dyed with alizarin red, scale bar = 1 mm. Figure B, Xenodermidae: Achalinus rufescens (BMNH 1983.193) completely everted and partially filled, scale bar = 2 mm; Fimbrios klossi (BMNH 1965.2.639) opened through a longitudinal slit, one lobe partially filled, scale bar = 1 mm; Pareidae: Pareas monticola (BMNH 1909.3.9.19) completely everted and filled, scale bar = 1 mm. Figure C, Pareidae: Asthenodipsas malaccanus (BMNH 1924.10.23.7) completely everted and filled; Aplopeltura boa (BMNH 94.6.30.63) completely everted and filled; scale bars = 2 mm. Figure D, Xylophiidae: Xylophis perroteti (BMNH 1955.1.3.10) opened through a longitudinal slit, spread flat, and dyed with alizarin red, scale bar = 5 mm. Figure E, Viperidae: Porthidum nasutum (MZUSP 7480) completely everted and filled; Vipera ammodytes (MZUSP 8223) completely everted and filled; scale bars = 5 mm. Figure F, Viperidae: Bothrops neuwiedi (MZUSP 11851) completely everted and filled; Causus bilineatus (MNHN 1993.5992) completely everted and filled; scale bars = 5 mm. Figure G, Homalopsidae: Homalopsis buccata (MNHN 1963.728) completely everted and filled; Brachyorrhos albus (FMNH 142324) completely everted and filled; scale bars = 5 mm. Figure H, Homalopsidae: Fordonia leucobalia (AMNH 107179) completely everted, filled, and dyed with alizarin red; Bitia hydroides (FMNH 229568) completely everted and filled; scale bars = 5 mm. Figure I, Homalopsidae: Erpeton tentaculatum (AMNH 8850) completely everted and filled, scale bar = 5 mm. Psammophiidae: Mimophis mahfalensis (UMMZ 209646) completely everted and partially filled, scale bar = 2 mm; Atractaspididae: Polemon christyi (FMNH 219912) completely everted and filled, scale bar = 5 mm. Figure J, Atractaspididae: Atractaspis fallax (AMNH 102298) completely everted and filled, scale bar = 10 mm; Macrelaps microlepidotus (FMNH 205863) completely everted and filled, scale bar = 5 mm. Figure K, 71 / 82 Cyclocoridae: Cyclocorus lineatus (MNHN 1900.411) opened through a longitudinal slit, spread flat, and dyed with alizarin red, scale bar = 5 mm; Oxyrhabdion modestum (FMNH 68907) completely everted, filled, and dyed with alizarin red, scale bar = 5 mm. Figure L, Lamprophiidae: Lamprophis fuliginosus (MNHN 1994.8111) completely everted and filled, scale bar = 5 mm; Chamaelycus fasciatum (BMNH 1909.4.29.2?3) completely everted and filled, scale bar = 2 mm; Lycodonomorphus rufulus (AMNH 140283) completely everted and filled, scale bar = 5 mm. Figure M, Lamprophiidae: Lycophidion semicinctus (MNHN 1995.3474) completely everted, filled, and dyed with alizarin red; Mehelya capensis (AMNH 73208) completely everted and filled; Pseudoboodon lemniscatus (MNHN 1905.185) completely everted and filled; scale bars = 5 mm. Figure N, Pseudoxyrhophiidae: Dromicodryas bernieri (UMMZ 218166) completely everted and filled, scale bar = 5 mm; Duberria lutrix (AMNH 115639) completely everted, filled, and dyed with alizarin red, scale bar = 3 mm; Alluaudina bellyi (UMMZ 209239) completely everted and filled, scale bar = 2 mm. Figure O, Pseudoxyrhophiidae: Pseudoxyrhopus tritaeniatus (UMMZ 195854) completely everted and filled; Liophidium torquatum (UMMZ 209430) completely everted and filled; scale bars = 5 mm. Figure P, Elapidae: Naja melanoleuca (BMNH 1959.1.7.69) completely everted and filled; Micrurus frontalis (IBSP 44331) completely everted and filled; scale bars = 5 mm. Figure Q, Elapidae: Austrelaps superbus (BMNH 1926.12.25.113) completely everted and filled; Bungarus candidus (BMNH 1937.11) completely everted and filled; scale bars = 5 mm. Figure R, Natricidae: Atretium schistosum (AMNH 85505) completely everted, filled, and dyed with alizarin red; Lycognathophis seychellensis (UMMZ 167994) completely everted, filled, and dyed with alizarin red; Afronatrix anoscopus (AMNH 142404) completely everted, filled, and dyed with alizarin red; scale bars = 2 mm. Figure S, Natricidae: Xenochrophis vittatus (BMNH 71.7.20.195?6) completely everted and filled; Natriciteres olivacea (AMNH 11905) completely everted and filled; Sinonatrix annularis (AMNH 84530) completely everted, filled, and dyed with alizarin red; scale bars = 2 mm. Figure T, Natricidae: Aspidura trachyprocta (AMNH 120248) completely everted and partially filled (no scale); Elapoidis fusca (MNHN 1895.55) completely everted and filled, scale bar = 2 mm. Figure U, Pseudoxenodontidae: Pseudoxenodon macrops (AMNH 34649) completely everted and filled; Dipsadidae: Conophis pulcher (MNHN 5981) completely everted and filled; scale bars = 5 mm. Figure V, Dipsadidae: Contia tenius (UMMZ 133370) completely everted and filled, scale bar = 2 mm; Urotheca decipiens (KU 103892) completely everted and filled, scale bar = 5 mm. Figure W, Dipsadidae: Oxyrhopus occipitalis (AMNH 129255) completely everted, filled, and dyed with alizarin red; Farancia erythrogramma (KU 197245) completely everted, filled, and dyed with alizarin red. Scale bars = 5 mm. Figure X, Dipsadidae: Tachymenis chilensis (MZUSP 8239) completely everted, filled, and dyed with alizarin red, scale bar = 2 mm; Heterodon nasicus (MNHN 3636) completely everted and filled, scale bar = 5 mm; Philodryas olfersii (IBSP 63455) completely everted, filled, and dyed with alizarin red, scale bar = 5 mm. Figure Y, Sibynophiidae: Sibynophis chinensis (AMNH 34102) completely everted and filled, scale bar = 2 mm; Scaphiodontophis annulatus (KU 191073) completely everted and filled, scale bar = 2 mm; Calamariidae: Pseudorabdion longiceps (BMNH 1969.1866) completely everted and partially filled, scale bar = 5 mm. Figure Z, Calamariidae: Calamaria lumbricoidis (BMNH 1928.2.18.26) completely everted and partially filled, scale bar = 5 mm; Calamaria linnaei (AMNH 31943) completely everted and partially filled, scale bar = 1 mm; Oreocalamus hanitschi (BMNH 1929.12.22.106) completely everted and partially filled, scale bar = 5 mm. Figure AA, Grayiidae: Grayia ornata (BMNH 98.3.25.3) completely everted and filled; Colubridae: Pantherophis guttatus (USNM 523605) completely everted and filled; Spilotes sulphureus (IBSP 68260) completely everted and filled. Scale bars = 10 mm. Figure AB, Colubridae: Dispholidus typus (AMNH 23110) completely everted and filled; Hierophis viridiflavus (MNHN 1978.414) completely everted and filled; 72 / 82 Boiga pulverulenta (MNHN 1967.437) completely everted and filled; scale bars = 10 mm. Figure AC, Colubridae: Ptyas korros (AMNH 84460) completely everted and filled, scale bar = 10 mm; Gongylosoma baliodeirus (MNHN 1989.199) completely everted and filled, scale bar = 3 mm; Liopeltis frenatus (MNHN 1928.75) completely everted and filled, scale bar = 5 mm. Figure AD, Elapoidea Incertae sedis: Buhoma depressiceps (MNHN 1991.1740) completely everted, partially filled, and dyed with alizarin red, scale bar = 1 mm. (PDF) Acknowledgments The authors wish to thank the following colleagues who kindly supplied tissue samples and/or allowed access to the specimens under their care: CW Myers, DR Frost, D Kizirian (AMNH); K de Queiroz, R McDiarmid, G Zug (USNM); A Dubois, A Ohler, I Ineich, R Bour (MNHN); P Campbell, D Gower (BMNH); MT Rodrigues (IBUSP); W Duellman, L Trueb (KU); D. Rossman, J. Boundy (LSUMZ); R. MacCulloch, A. Lathrop (ROM); M-O Ro?del, F Tillack, R Gu?nther (ZMB); A Resetar (FMNH); R Nussbaum, G Schneider (UMMZ); G Puorto, FL Franco (IBSP); B Schacham, Y Werner (HUJR). We are especially indebted to AB Carvalho and R Rodrigues for scanning specimens and providing technical support for figure preparations, to L. Oliveira for helping with the scanning electron microscopy of maxillary teeth, and to T. Rowe and J. Maisano for providing 3D reconstructions of squamate taxa scanned as part of the DigiMorph project. We are grateful to C Sarturi and A Schnorr (PUCRS) and Dr. Y Gao, P-T Luan and J-X Wang (KIZ) for laboratory assistance. FGG. and RG were supported by scholarships from Fundac??o de Amparo ? Pesquisa do Estado de S?o Paulo (FAPESP grant numbers 2007/52781-5, 2012/ 08661?3, 2007/52144-5, 2011/2167-4, 2008/52285-0, 2012/ 24755-8, 2016/13469-5). Funding for this study was provided by FAPESP (2002/13602-4, 2011/50206-9 and 2016/50127-5). Conceptualization: Hussam Zaher, Robert W. Murphy, Felipe G. Grazziotin. Data curation: Hussam Zaher, Roberta Graboski, Kristin Mahlow, Felipe G. Grazziotin. Formal analysis: Hussam Zaher, Roberta Graboski, Mark Wilkinson, Felipe G. Grazziotin. Funding acquisition: Hussam Zaher. Investigation: Hussam Zaher, Mark Wilkinson, Felipe G. Grazziotin. Methodology: Hussam Zaher, Mark Wilkinson, Felipe G. Grazziotin. Project administration: Hussam Zaher, Ya-Ping Zhang. Resources: Hussam Zaher, Robert W. Murphy, Ya-Ping Zhang, Felipe G. Grazziotin. Software: Felipe G. Grazziotin. Supervision: Hussam Zaher, Robert W. Murphy, Ya-Ping Zhang, Felipe G. Grazziotin. Validation: Robert W. Murphy, Juan Camilo Arredondo, Roberta Graboski, Paulo Roberto Machado-Filho, Giovanna G. Montingelli, Ana Bottallo Quadros, Nikolai L. Orlov, Mark Wilkinson, Ya-Ping Zhang, Felipe G. Grazziotin. Visualization: Juan Camilo Arredondo, Roberta Graboski, Paulo Roberto Machado-Filho, Kristin Mahlow, Giovanna G. Montingelli, Ana Bottallo Quadros, Nikolai L. Orlov, Mark Wilkinson, Ya-Ping Zhang, Felipe G. Grazziotin. 73 / 82 Writing ? original draft: Hussam Zaher, Juan Camilo Arredondo, Roberta Graboski, Paulo Roberto Machado-Filho, Giovanna G. Montingelli, Ana Bottallo Quadros, Mark Wilkinson, Felipe G. Grazziotin. Writing ? review & editing: Robert W. Murphy, Juan Camilo Arredondo, Roberta Graboski, Paulo Roberto Machado-Filho, Kristin Mahlow, Giovanna G. Montingelli, Ana Bottallo Quadros, Nikolai L. Orlov, Mark Wilkinson, Ya-Ping Zhang, Felipe G. Grazziotin. 1. Underwood G. A contribution to the classification of snakes. London: British Museum of Natural History; 1967. 3. Haas G. Remarques concernant les relations phyloge?niques des diverses familles d?ophidiens fonde?es sur la diffe?renciation de la musculature mandibulaire. Colloq Int Cent Nat Rech Sci. 1962; 104: 215?241. 74 / 82 Wallach V, Williams KL, Boundy J. Snakes of the World: A Catalogue of Living and Extinct Species. Boca Raton: Taylor & Francis; 2014. 35. Savage JA. What are the correct family names for the taxa that include the snake genera Xenodermus, Pareas, and Calamaria. Herpetol Rev. 2015; 46: 664?665. Weinell JL, Brown RM. Discovery of an old, archipelago-wide, endemic radiation of Philippine snakes. Mol phylogenet evol. 2018; 119: 144?150. https://doi.org/10.1016/j.ympev.2017.11.004 PMID: 29162550 Wilkinson M. Majority-rule reduced consensus trees and their use in bootstrapping. Mol Biol evol. 1996; 13: 437?444. https://doi.org/10.1093/oxfordjournals.molbev.a025604 PMID: 8742632 38. Rhodin AG, Kaiser H, van Dijk PP, Wu?ster W, O?Shea M, Archer M, et al. Comment on Spracklandus Hoser, 2009 (Reptilia, Serpentes, Elapidae): request for confirmation of availability of the generic name and for the nomenclatural validation of the journal in which it was published (Case 3601; BZN 70: 234?237; 71: 30?38; 133?135,181?182, 252?253). Bull Zool Nomencl. 2015; 72: 65?78. 39. Federhen S, Hotton C, Mizrachi I. Comments on the paper by Pleijel et al. (2008): vouching for GenBank. Mol Phylogenet Evol. 2009; 53: 357?358. https://doi.org/10.1016/j.ympev.2009.04.016 PMID: 19410006 40. Bricker J, Bushar LM, Reinert HK, Gelbert L. Purification of high quality DNA from shed skin. Herpetol. Rev. 1996; 27: 133?134. 41. Hillis DM, Mable BK, Moritz C. Applications of molecular systematics. In: Hillis DM, Moritz C, Mable BK, editors. Molecular Systematics. Sunderland, MA: Sinauer Associates Inc; 1996. pp. 515?544. 42. Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, et al. Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics. 2012; 28: 1647?1649. https://doi.org/10.1093/bioinformatics/bts199 PMID: 22543367 43. Katoh K, Misawa K, Kuma K, Miyata T. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 2002; 30: 3059?3066. https://doi.org/10.1093/nar/ gkf436 PMID: 12136088 75 / 82 44. Wiens JJ. Missing data, incomplete taxa, and phylogenetic accuracy. Syst Biol. 2003; 52: 528?538. PMID: 12857643 47. Stamatakis A. RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics. 2006; 22: 2688?2690. https://doi.org/10.1093/bioinformatics/ btl446 PMID: 16928733 53. Yang Z. PAML 4: phylogenetic analysis by maximum likelihood. Mol Biol Evol. 2007; 24: 1586?1591. https://doi.org/10.1093/molbev/msm088 PMID: 17483113 54. Sanderson M.J. Estimating absolute rates of molecular evolution and divergence times: a penalized likelihood approach, Mol. Biol. Evol. 2002: 19: 101?109. https://doi.org/10.1093/oxfordjournals. molbev.a003974 PMID: 11752195 55. Smith SA, O?meara BC. treePL: divergence time estimation using penalized likelihood for large phylogenies. Bioinformatics. 2012: 2689?2690. https://doi.org/10.1093/bioinformatics/bts492 PMID: 22908216 56. Ho SYW, Duch?ne S. Molecular-clock methods for estimating evolutionary rates and timescales. Mol. Ecol. 2014; 23: 5947?5965. https://doi.org/10.1111/mec.12953 PMID: 25290107 57. Bell CD. Between a rock and a hard place: applications of the ?molecular clock? in systematic biology. Syst. Bot. 2015; 40: 6?13. 58. Hsiang AY, Field DJ, Webster TH, Behlike ADB, Davis MB, Racicot, et al. The origin of snakes: revealing the ecology, behavior, and evolutionary history of early snakes using genomics, phenomics, and the fossil record. BMC Evol Biol. 2015; 15: 1?22. 59. Chalifa Y, Tchernov E. Pachyamia latimaxillaris, new genus and species (Actinopterygii: Amiidae), from the Cenomanian of Jerusalem. J Vert Paleontol. 1982; 2: 269?285. 60. Chalifa Y. Saurorhamphus judeaensis (Salmoniformes: Enchodontidae), a new longirostrine fish from the Cretaceous (Cenomanian) of Ein-Yabrud, near Jerusalem. J Vert Paleontol. 1985; 5: 181?193. 61. Head JJ. Fossil calibration dates for molecular phylogenetic analysis of snakes 1: Serpentes, Alethinophidia, Boidae, Pythonidae. Palaeo Electronica. 2015; 18: 1?17. 62. Ogg JG, Hinnov LA. The Cretaceous Period. In: Gradstein FM, Ogg JG, Schmitz MD, Ogg GM, editors. The Geologic Time Scale. New York: Elsevier Publishing Company; 2012. pp. 793?854. 63. Head JJ, Bloch JI, Hastings AK, Bourque JR, Cadena EA, Herrera FA, et al. Giant boid snake from the Paleocene neotropics reveals hotter past equatorial temperatures. Nature. 2009; 457: 715?717. https://doi.org/10.1038/nature07671 PMID: 19194448 64. Albino AM, Brizuela S. An overview of the South American fossil squamates. Anat Rec. 2014; 297: 349?368. 65. Head JJ, Bloch JI, Moreno-Bernal J, Rincon Burbano A, Bourque J. Cranial osteology, body size, systematics, and ecology of the giant Paleocene snake Titanoboa cerrejonensis. J Vert Paleontol. 2013; 33: 140?141. 66. Jaramillo CA, Mu?oz F, Cogollo M, De La Parra F. Quantitative biostratigraphy for the Paleocene of the Llanos foothills, Colombia: improving palynological resolution for oil exploration. In: Powell AJ, Riding JB, editors. Recent Developments in Applied Biostratigraphy. The Micropalaeontological Society, Special Publication; 2005. pp. 145?159. 76 / 82 67. Jaramillo C, Bayona G, Pardo-Trujillo AP, Rueda M, Torres V, Harrington GJ, Mora G. The palynology of the Cerrej o?n Formation (upper Paleocene) of northern Colombia. Palynology. 2007; 31: 153? 189. 68. Rage JC, Folie A, Rana RS, Singh H, Rose KD, Smith T. A diverse snake fauna from the early Eocene of Vastan Lignite Mine, Gujarat, India. Acta Palaeontol Pol. 2008; 53: 391?403. 77. Clementz M, Bajpai S, Ravikant V, Thewissen JGM, Singh IB, Prasad V. Early Eocene warming events and the timing of terrestrial faunal exchange between India and Asia. Geology. 2011; 39: 15? 18. 78. Szyndlar Z, Bo?hme W. Die fossilen Schlangen Deutschlands Geschichte der faunen und ihrer Erforschung. Mertensiella. 1993; 3:381?432. 79. Szyndlar Z, Rage JC. Oldest fossil vipers (Serpentes, Viperidae) from the Old World. In: Joger U, Editor. Phylogeny and systematics of the Viperidae. Kaupia, Darmsta?dter Beitra?ge zur Naturgeschichte; 1999. pp. 9?20. 80. Dowling HG, Duellman WE. Systematic herpetology: a synopsis of families and higher categories. New York: HISS Publications; 1978. 81. Schleich HH, Ka?stle W, Kabisch K. Amphibians and Reptiles of North Africa. Koeltz, Koenigstein; 1996. 82. Saint Girons H. Bioge?ographie et e?volution des vip?res Europe?ennes. C. R. Soc. Bioge?ogr. 1980; 496: 146?172. 83. Nilson G. and Andren C. Evolution, systematics and biogeography of Palaearctic vipers. Symp. Zool. Soc. 1997; 70: 31?42. 84. Herrmann HW, Joger U. Evolution of viperine snakes. In: Thorpe RS, Wuster W, Malhotra A, editors. Venomous snakes: Ecology, evolution and snakebite. Oxford, Symp Zool Soc London; 1997. pp. 43? 61. 85. Szyndlar Z, Rage JC. Fossil record of the true vipers. In: Schuett G, Ho?ggren M, Douglas ME, Greene HW, editors. Biology of the vipers. Eagle Mountain Publishing, Eagle Mountain; 2002. pp. 419?444. 86. ?er?ansky? A, Rage JC, Klembara J. The Early Miocene squamates of Amo?neburg (Germany): the first stages of modern squamates in Europe. J Syst Palaeontol. 2015; 13: 97?128. 87. Bo?hme, M., & Ilg, A. 2003 [cited 2018 Jul 17]. Database of Vertebrates: fossil Fishes. Amphibians, Reptiles and Birds (fosFAR-base) localities and taxa from the Triassic to the Neogene. www.wahrestaerke.com. 88. Ivanov M. The first European pit viper from the Miocene of Ukraine. Acta Paleontol Pol. 1999; 44: 327?334. 89. Holman JA. A herpetofauna from an eastern extension of the Harrison Formation early Miocene Arikareean Cherry County Nebraska. J Vertebr Paleontol. 1981; 1: 49?56. 90. Holman JA, Tanimoto M. cf. Trimeresurus Lace?p?de (Reptilia: Squamata: Viperidae: Crotalinae) from the late Early Miocene of Japan. Acta Zool. Cracov. 2004; 47: 1?7. 77 / 82 91. Agust?? J, Cabrera L, Garce?s M, Pare?s JM. The Vallesian mammal succession in the Vall?s-Pened?s basin (NE Spain): Paleomagnetic calibration and correlation with global events. Palaeogeogr Palaeoclimatol Palaeoecol. 1997; 133: 3?4. 98. Stevens NJ, O?Connor PM, Roberts EM, Gottfried MD. A hyracoid from the Late Oligocene Red Sandstone Group of Tanzania, Rukwalorax jinokitana (gen. and sp. nov.). J Vert Paleontol. 2009; 29: 972? 975 99. Stevens NJ, Holroyd PA, Roberts EM, O?Connor PM, Gottfried MD. Kahawamys mbeyaensis (n. gen. n. sp.) (Rodentia: Thryonomyoidea) from the Late Oligocene Rukwa Rift Basin, Tanzania. J Vert Paleontol. 2009; 29: 631?634. 100. Roberts EM, O?Connor PM, Gottfried MD, Stevens NJ, Kapilima S, Ngsala S. Revised stratigraphy and age of the Red Sandstone Group in the Rukwa Rift Basin, Tanzania. Cret Res. 2004; 25: 749? 759. 101. Roberts EM, O?Connor PM, Stevens NJ, Gottfried MD, Jinnah ZA, et al. Sedimentology and depositional environments of the Red Sandstone Group, Rukwa Rift Basin, southwestern Tanzania: New insight into Cretaceous and Paleogene terrestrial ecosystems and tectonics in sub-equatorial Africa. J Afr Earth Sci. 2010; 57: 179?212. 102. Roberts EM, Stevens NJ, O?Connor PM, Dirks PHGM, Gottfried MD, et al. Initiation of the western branch of the East African Rift coeval with the eastern branch. Nat Geosci. 2010; 5: 289?294. 103. Scanlon JD, Lee MS, Archer M. Mid-Tertiary elapid snakes (Squamata, Colubroidea) from Riversleigh, northern Australia: early steps in a continent-wide adaptive radiation. Geobios. 2003; 36: 573?601. 104. Sanders KL, Lee MSY. Molecular evidence for a rapid late-Miocene radiation of Australasian venomous snakes (Elapidae, Colubroidea). Mol. Phylogenet. Evol. 2008; 46: 1165?1173. https://doi.org/10. 1016/j.ympev.2007.11.013 PMID: 18222097 105. Travouillon KJ, Archer M, Hand SJ, Godthelp H. Multivariate analyses of Cenozoic mammalian faunas from Riversleigh, northwestern Queensland. Alcheringa. 2006; 30: 323?349 106. Arena KR, Hartmann LA, Lana CL. Evolution of neoproterozoic ophilites from the southern Brasiliano Orogeny revealed by zircon U-Pb-Hf isotopes and geochemistry. Precambr Res. 2016; 203: 1?12. 107. Archer M, Hand SJ, Godthelp H, Creaser P. Correlation of the Caenozoic sediments of the Riversleigh World Heritage fossil property, Queensland, Australia. Me?m. Trav. Inst. Montpellier E?cole Prat. Hautes E? tudes. 1997; 21: 131?152. 108. Myers TJ, Crosby K, Archer M, Tyler M. The Encore Local Fauna, a late Miocene assemblage from Riversleigh, northwestern Queensland. Mem Assoc Australasian Palaeontol. 2001; 25: 147?154. 109. Smith K. New constraints on the evolution of the snake clades Ungaliophiinae, Loxocemidae and Colubridae (Serpentes), with comments on the fossil history of erycine boids in North America. Zool Anz. 2013; 252: 157?182. Rage JC. Les serpents des phosphorites du Quercy. Palaeovertebrata. 1974; 6: 274?303. Hoganson JW, Murphy EC, Forsman NF. Lithostratigraphy, paleontology, and biochronology of the Chadron, Brule, and Arikaree Formations in North Dakota, In: Terry DO, LaGarry HE, Hunt RM, editors. Depositional Environments, Lithostratigraphy, and Biostratigraphy of the White River and Arikaree Groups (Late Eocene to Early Miocene, North America) (Geological Society of America Special Paper 325). Colorado: Geological Society of America, Boulder; 1998. pp. 185?196. 112. Pearson DA, Hoganson JW. The Medicine Pole Hills local fauna: Chadron Formation (Eocene: Chadronian), Bowman County, North Dakota. Proc North Dakota Acad Sci. 1995; 49: 115?121. 78 / 82 113. Smith KT. The evolution of mid-latitude faunas during the Eocene: Late Eocene lizards of the Medicine Pole Hills reconsidered. Bull Peabody Mus Nat Hist. 2011; 52: 3?105. 114. Rage JC, Szyndlar Z. Natrix longivertebrata from the European Neogene, a snake with one of the longest known stratigraphic ranges. Neues Jahrb Geol P-M. 1986; 1: 56?64. 116. Pokrant F, Kindler C, Ivanov M, Cheylan M, Geniez P, Bo?hme W, Fritz U. Integrative taxonomy provides evidence for the species status of the Ibero-Maghrebian grass snake Natrix astreptophora. Biol J Linn Soc. 2016; 118: 873?888. 117. Rage JC. The oldest known colubrid snakes. The state of the art. Acta Zool. Cracov. 1988; 31: 457? 474. 118. Ivanov M. The oldest known Miocene snake fauna from Central Europe: Merkur-North locality, Czech Republic. Acta Palaeontol Pol. 2002; 47: 513?534. 119. Holman J.A. Upper Miocene snakes (Reptilia, Serpentes) from Southeastern Nebraska. J Herpetol. 1997; 11: 323?335. 120. Cundall D, Rossman DA. Quantitative comparisons of skull form in the colubrid snake genera Farancia and Pseudoeryx. Herpetologica. 1984; 40: 388?405. 121. Tedford RH, Albright B III, Barnosky AD, Ferrusquia-Villafranca I, Hunt RM jr, Storer JE, Swisher CC III, Voorhies MR, Webb SD, Whistler DP. Mammalian Biochronology of the Arikareean Through Hemphillian Interval (Late Oligocene Through Early Pliocene Epochs). In Woodburne MO, editor. Late Cretaceous and Cenozoic Mammals of North America: Biostratigraphy and Geochronology. New York, Columbia University Press; 2004. pp. 169?231. 122. Bair AR. Description of a new species of the North American archaic pika Hesperolagomys (Lagomorpha: Ochotonidae) from the middle Miocene (Barstovian) of Nebraska and reassessment of the genus Hesperolagomys. Palaeontol Electronica. 2011; 14: 49. 123. Hilgen FJ, Lourens LJ, Van Dam JA. The Neogene Period, In Gradstein F, Ogg JG, Schmitz M, Ogg G, editors. The Geologic Time Scale 2012. Elsevier; 2012. pp. 923?978. 124. Alencar LR, Quental TB, Grazziotin FG, Alfaro ML, Martins M, Venzon M, Zaher H. Diversification in vipers: Phylogenetic relationships, time of divergence and shifts in speciation rates. Mol Phylogenet Evol. 2016; 105: 50?62. https://doi.org/10.1016/j.ympev.2016.07.029 PMID: 27480810 125. Underwood G., Kochva E. On the affinities of the burrowing asps Atractaspis (Serpentes: Atractaspididae). Zool J Linn Soc. 1993; 107: 3?64. 126. Moyer K, Jackson K. Phylogenetic relationships among the Stiletto Snakes (genus Atractaspis) based on external morphology. Afr J Herpetol. 2011; 60: 30?46. 127. Pyron RA, Guayasamin JM, Pe?afiel N, Bustamante L, Arteaga A. Systematics of Nothopsini (Serpentes, Dipsadidae), with a new species of Synophis from the Pacific Andean slopes of southwestern Ecuador. ZooKeys. 2015; 541: 109?147. 128. Wang X, Messenger K, Zhao E, Zhu C. Reclassification of Oligodon ningshaanensis Yuan, 1983 (Ophidia: Colubridae) into a new genus, Stichophanes gen. nov. with description on its malacophagous behavior. Asian Herpetol Res. 2014; 5: 137?149. 129. Cadle JE. Molecular systematics of Neotropical xenodontine snakes. I. South American xenodontines. Herpetologica. 1984; 40: 8?20. 130. Cadle JE. Molecular systematics of Neotropical xenodontine snakes. II. Central American xenodontines. Herpetologica. 1984; 40: 21?30. 131. Pyron RA, Arteaga A, Echevarria LY, Torres-Carvajal O. A revision and key for the tribe Diaphorolepidini (Serpentes: Dipsadidae) and checklist for the genus Synophis. Zootaxa. 2016; 4171: 293?320. https://doi.org/10.11646/zootaxa.4171.2.4 PMID: 27701225 132. Burbrink F, Pyron RA. The taming of the skew: estimating proper confidence intervals for divergence dates. Syst. Biol. 2008; 57: 317?328. https://doi.org/10.1080/10635150802040605 PMID: 18432552 133. Wu?ster W, Peppin L, Pook CE, Walker DE. A nesting of vipers: phylogeny and historical biogeography of the Viperidae (Squamata: Serpentes). Molecular Phylogenetics and Evolution. 2008; 49: 445?459. https://doi.org/10.1016/j.ympev.2008.08.019 PMID: 18804544 134. Pyron RA, Burbrink F. Extinction, ecological opportunity, and the origins of global snake diversiy. Evolution. 2012; 66: 163?178. https://doi.org/10.1111/j.1558-5646.2011.01437.x PMID: 22220872 135. Vidal N, Rage JC, Couloux A, Hedges SB. Snakes (Serpentes). In: Hedges SB, Kumar S, editors. The Timetree of life. New York: Oxford University Press; 2009. pp. 390?397. 136. Streicher JW, Wiens JJ. Phylogenomic analyses reveal novel relationships among snake families. Mol Phylogenet Evol. 2016; 100: 160?169. https://doi.org/10.1016/j.ympev.2016.04.015 PMID: 27083862 79 / 82 137. Pyron RA, Hendry CR, Chou VM, Lemmon EM, Lemmon AR, Burbrink FT. Effectiveness of phylogenomic data and coalescent species-tree methods for resolving difficult nodes in the phylogeny of advanced snakes (Serpentes: Caenophidia). Mol Phylogenet Evol. 2014; 81: 221?231. https://doi. org/10.1016/j.ympev.2014.08.023 PMID: 25193610 138. Ikeda T. A comparative morphological study of the vertebrae of snakes occurring in Japan and adjacent regions. Curr Herpetol. 2007; 26: 13?34. 139. Rieppel O. The performance of morphological characters in broad-scale phylogenetic analyses. Biol J Linn Soc Lond. 2007; 92: 297?308. 140. Rieppel O, Zaher H. The development of the skull in Acrochordus granulatus (Schneider) (Reptilia: Serpentes), with special consideration of the otico-occipital complex. J Morphol. 2001; 249: 252?266. https://doi.org/10.1002/jmor.1053 PMID: 11517468 142. Gauthier JA, Kearney M, Maisano JA, Rieppel O, Behlke AD. Assembling the squamate tree of life: perspectives from the phenotype and the fossil record. Bull Peabody Mus Nat Hist. 2012; 53: 3?308. 143. Dowling HG, Maha GC, Maxson LR. Biochemical evaluation of colubrid snake phylogeny. J Zool. 1983; 201: 309?329. 144. Sharma GP, Nakhasi U. Chromosomal polymorphism in three species of Indian snakes. Cytobios. 1979; 24: 167?179. PMID: 527373 145. Sharma GP, Nakhasi U. Karyological studies on six species of Indian snakes (Colubridae: Reptilia). Cytobios. 1980; 27: 177?192. PMID: 7428439 146. Rovatsos M, Vukic J, Lymberakis P, Kratochv??l L. Evolutionary stability of sex chromosomes in snakes. Proc R Soc B. 2015; 282: 20151992. https://doi.org/10.1098/rspb.2015.1992 PMID: 26702042 147. Teynie? A, David P, Lottier A, Le MD, Vidal N, Nguyen TQ. A new genus and species of xenodermatid snake (Squamata: Caenophidia: Xenodermatidae) from northern Lao People?s Democratic Republic. Zootaxa. 2015; 3926: 523?540. https://doi.org/10.11646/zootaxa.3926.4.4 PMID: 25781800 148. Hardaway TE, Williams KL. Costal cartilages in snakes and their phylogenetic significance. Herpetologica. 1976: 378?387. 149. Persky B, Smith HM, Williams KL. Additional observations on ophidian costal cartilages. Herpetologica. 1976; 32: 399?401. 150. Underwood G. A comprehensive approach to the classification of higher snakes. Herpetologica. 1967; 23: 161?168. 151. Sim?es BF, Sampaio FL, Douglas RH, Kodandaramaiah U, Casewell NR, Harrison RA, Hart NS, Partridge JC, Hunt DM, Gower DJ. Visual Pigments, Ocular Filters and the Evolution of Snake Vision. Mol Biol Evol. 2016; 33: 2483?2495. https://doi.org/10.1093/molbev/msw148 PMID: 27535583 152. Ruane S, Austin CC. Phylogenomics using formalin-fixed and 100+ year-old intractable natural history specimens. Mol Ecol Resour. 2017; 17: 1003?1008. https://doi.org/10.1111/1755-0998.12655 PMID: 28118515 153. Deepak V, Ruane S, Gower DJ. A new subfamily of fossorial colubroid sankes from the western Ghats of peninsular India. J Nat Hist. 2019; 52: 2919?2934. 154. Underwood G. An overview of venomous snake evolution. In: Thorpe ES, Wu?ster W, Malhorta A, editors. Venomous snakes: ecology, evolution and snakebite. New York: Oxford University Press; 1997. pp. 1?13. 155. Vonk FJ, Admiraal JF, Jackson K, Reshef R, de Bakker MAG et al. Evolutionary origin and development of snake fangs. Nature. 2008; 454: 630?633. https://doi.org/10.1038/nature07178 PMID: 18668106 156. Nagy ZT, Gvoz?d??v V, Meirte D, Collet M, Pauwels OS. New data on the morphology and distribution of the enigmatic Schouteden?s sun snake, Helophis schoutedeni (de Witte, 1922) from the Congo Basin. Zootaxa. 2014; 3755: 96?100. https://doi.org/10.11646/zootaxa.3755.1.5 PMID: 24869811 157. Smith MA. The Fauna of British India, Ceylon and Burma, Including the Whole of the Indo-Chinese Sub-Region. Reptilia and Amphibia. 3 (Serpentes). Taylor and Francis; 1943. 158. Bourquin O. A new genus and species of snake from the Natal Drakensberg, South Africa. Annals Transvaal Mus. 1991; 35: 199?203. 159. Lambiris AJL. Description of the hemipenis of Montaspis gilvomaculata Bourquin, 1991 (Serpentes, Colubridae). Annals of the Natal Museum. 1997; 38: 1?3. 160. Boulenger GA. Catalogue of the Snakes in the British Museum (Natural History). Volume II, Containing the Conclusion of the Colubridae Aglyphae. Taylor and Francis; 1894. 80 / 82 161. Leviton AE. Contributions to a review of Philippine snakes, II. The snakes of the genera Liopeltis and Sybinophis. Philipp J. Sci. 1963; 92: 367?381. 162. Boulenger G.A. Catalogue of the snakes in the British Museum (Nat. Hist.). Volume I. Taylor and Francis; 1893. 163. Orlov NL, Kharin VE, Anajeva NB, Tao NT, Truong NQ. A new genus and species of colubrid snake (Squamata, Ophidia, Colubridae) from south Vietnam. Russian J Herpetol. 2009; 16: 228?240. 164. Jackson K, Underwod G, Arnold EN, Savitzky AH. 1999. Hinged teeth in the enigmatic colubrid, Iguanognathus werneri. Copeia. 1999; 3: 815?818. 165. Boulenger GA. Description of a new genus of aglyphous colubrine snakes from Sumatra. Ann. Mag. Nat. Hist. 1898; 7: 73. 166. Tonini JFR, Beard KH, Ferreira RB, Jetz W, & Pyron RA. Fully-sampled phylogenies of squamates reveal evolutionary patterns in threat status. Biological Conservation. 2016; 204: 23?31. 167. Bell CD. Between a rock and a hard place: Applications of the ?molecular clock? in systematic biology. Syst Bot. 2015; 40: 6?13. 168. Rage JC, Werner C. Mid-Cretaceous (Cenomanian) snakes from Wadi Abu Hashim, Sudan: the earliest snake assemblage. Palaeontol Afr. 1999; 35: 85?110. 169. Head JJ. Snakes of the Siwalik group (Miocene of Pakistan): systematics and relationship to environmental change. Palaeo Electronica. 2005; 33: 8.1. 18A. 170. Rage JC, Ford RLE. Amphibians and squamates from the Upper Eocene of the Isle of Wight. Tertiary Res. 1980; 3: 47?60. 171. Parmley D Holman JA. Nebraskophis Holman from the late Eocene of Georgia (USA), the oldest known North American colubrid snake. Acta Zoologica Cracoviensia. 2003; 46: 1?8. 172. McCartney JA, Seiffert ER. A Late Eocene snake fauna from the Fayum depression, Egypt. J. Vert. Paleontol. 2015; 36: e1029580. 173. Holman JA. A new genus of primitive colubroid snake from the Upper Eocene, Isle of Wight, England. Tertiary Res. 1993; 14: 151?154. 174. Rage JC, Pickford M, Senut B. Amphibians and squamates from the middle Eocene of Namibia, with comments on pre-Miocene anurans from Africa. Ann Paleontol. 2013; 99: 217?242. 175. Rage JC, Buffetaut E, Buffetaut-Tong H, Chaimanee Y, Ducrocq S, Jaeger JJ, Suteethorn V. A colubrid snake in the late Eocene of Thailand: the oldest known Colubridae (Reptilia, Serpentes). C R Acad Sci II. 1992; 314: 1085?1089. 176. Prothero DR. The Eocene-Oligocene transition: paradise lost. New York: Columbia University Press; 1994. 177. Costa E, Garce?s M, Sa?ez A, Cabrera L, Lo?pez-Blanco M. The age of the ?Grande Coupure? mammal turnover: New constraints from the Eocene?Oligocene record of the Eastern Ebro Basin (NE Spain), Palaeogeogr Palaeoclimatol Palaeoecol. 2011; 301: 97?107. 178. Holman J. A. Fossil Snakes of North America: Origin, Evolution, Distribution, Paleoecology. Indiana University Press, Bloomington, Indiana; 2000. 179. Szyndlar Z. Early Oligocene to Pliocene Colubridae of Europe: a review. Bull Soc geol Fr. 2012; 183: 661?681. Prothero DR. The late Eocene-Oligocene extintions. Annu Rev Earth Planet Sci. 1994; 22: 145?165. Zanazzi A, Kohn MJ, MacFadden BJ, Terry DO. Large temperature drop across the Eocene-Oligocene transition in central North America. Nature. 2007; 445: 639?642. https://doi.org/10.1038/ nature05551 PMID: 17287808 182. Pearson PN, McMillan IK, Wade BS, Jones TD, Coxall HK, Bown PR, Lear CH. Extintion and environmental change across the Eocene-Oligocene boundary in Tanzania. Geology. 2008; 36: 179?182. 183. Liu Z, Pagani M, Zinniker D, DeConto R, Huber M, Brinkhuis, et al. Global cooling during the EoceneOligocene climate transition. Science. 2009; 323: 1187?1190. https://doi.org/10.1126/science. 1166368 PMID: 19251622 Kinkelin F. Ein fossiler Giftzahn. Zool. Anz. 1892; 15: 93?94. Rage JC, Ginsburg L. Amphibians and squamates from the early Miocene of Li Mae Long, Thailand: The richest and most diverse herpetofauna from the Cainozoic of Asia. In: Rocek Z, Hart S. editors. Herpetology?97; 1997. pp. 167?168. 186. Holman JA. A herpetofauna from an eastern extension of the Harrison Formation (Early Miocene: Arikareean), Cherry County, Nebraska. J Vert Paleontol. 1981; 1: 49?56. 187. Chkhikvadze VM. Preliminary results of the study of Tertiary amphibians and squamate reptiles of the Zaisan Basin. In: 6th All-Union Herpetol. Conference, Tashkent; 1985. pp. 234?235. 81 / 82 Bailon S. Les amphibiens et les reptiles du Plioc?ne superieur de Balaruc II. Palaeovertebrata. 1989; 19: 7?28. 2. Hoffstetter R . Contribution ? l'e?tude des Elapidae actuels et fossils et de l' oste?ologie des ophidiens . Arch Mus Hist Nat. Lyon . 1939 ; 15 : 1 - 75 . 4. Hoffstetter R , Gayrard Y. Observations sur l'oste?ologie et la classification des Acrochordidae (Serpentes) . Bull Mus Natl Hist Nat . 1965 ; 36 : 677 - 696 . 5. Romer AS . Osteology of the reptiles . Chicago: University of Chicago Press; 1956 . 6. Groombridge B . On the vomer in Acrochordidae (Reptilia: Serpentes), and its cladistic significance . J Zool . 1979 ; 189 : 559 - 567 . 7. Groombridge B. Variations in morphology of the superficial palate of henophidian snakes and some possible systematic implications . J Nat Hist . 1979 ; 13 : 447 - 475 . 8. McDowell SB. Systematics . In: Seigel RA , Collins JTC , Novak SS , editors. Snakes: Ecology and Evolutionary Biology . New York: MacMillan; 1987 . pp. 1 - 50 . 9. Rieppel O. A review of the origin of snakes . Evol Biol . 1988 ; 22 : 37 - 130 . 10. Rieppel O , Zaher H. The development of the skull in Acrochordus granulatus (Schneider) (Reptilia: Serpentes), with special consideration of the otico-occipital complex . J Morphol . 2001 ; 249 : 252 - 266 . https://doi.org/10.1002/jmor.1053 PMID: 11517468 11. Bourgeois M. Contribution ? la morphologie compare?e du cra?ne des ophidiens de l'Afrique Centrale . Publ Univ Off Congo . 1968 ; 18 : 1 - 295 . 12. McDowell SB . Affinities of the snakes usually called Elaps lacteus and E. dorsalis. Zool J Linn Soc . 1968 ; 47 : 561 - 578 . 13. Kraus F , Brown WM . Phylogenetic relationships of colubroid snakes based on mitochondrial DNA sequences . Zool J Linn Soc . 1998 ; 122 : 455 - 487 . 14. Keogh JS . Molecular phylogeny of elapid snakes and a consideration of their biogeographic history . Biol J Linn Soc . 1998 ; 63 : 177 - 203 . 15. Gravlund P. Radiation within the advanced snakes (Caenophidia) with special emphasis on African opistoglyph colubrids, based on mitochondrial sequence data . Biol J Linn Soc . 2001 ; 72 : 99 - 114 . 16. Vidal N , Hedges SB . Higher-level relationships of caenophidian snakes inferred from four nuclear and mitochondrial genes . C R Acad Sci Paris, Biol. 2002 ; 325 : 987 - 995 . 17. Kelly CMR , Barker NP , Villet MH . Phylogenetics of advanced snakes (Caenophidia) based on four mitochondrial genes . Syst Biol . 2003 ; 52 : 439 - 459 . PMID: 12857637 18. Lawson R , Slowinski JB , Crother BI , Burbrink FT . Phylogeny of the Colubroidea (Serpentes): new evidence from mitochondrial and nuclear genes . Mol Phylogenet Evol . 2005 ; 37 : 581 - 601 . https://doi. org/10.1016/j.ympev. 2005 . 07 .016 PMID: 16172004 19. Vidal N , Delmas AS , David P , Cruaud C , Couloux A , Hedges SB . The phylogeny and classification of caenophidian snakes inferred from seven nuclear protein-coding genes . C R Biol . 2007 ; 330 : 182 - 187 . https://doi.org/10.1016/j.crvi. 2006 . 10 .001 PMID: 17303545 20. Vidal N , Branch WR , Pauwels OSG , Hedges SB , Broadley DG , Wink M , et al. Dissecting the major African snake radiation: a molecular phylogeny of the Lamprophiidae Fitzinger (Serpentes, Caenophidia) . Zootaxa . 2008 ; 1945 : 51 - 66 . 21. Grazziotin FG , Zaher H , Murphy RW , Scrocchi G , Benavides MA , Zhang YP , et al. Molecular phylogeny of the New World Dipsadidae (Serpentes: Colubroidea): a reappraisal . Cladistics . 2012 ; 28 : 437459 . 22. Kelly CMR , Barker NP , Villet MH , Broadley DG . Phylogeny, biogeography and classification of the snake superfamily Elapoidea: a rapid radiation in the late Eocene . Cladistics, 2009 ; 25 : 38 - 63 . 23. Zaher H , Grazziotin FG , Cadle JE , Murphy RW , de Moura JC , Bonatto SL . Molecular phylogeny of advanced snakes (Serpentes, Caenophidia) with an emphasis on South American xenodontines: a revised classification and descriptions of new taxa . Pap Avulsos Zool . 2009 ; 49 : 115 - 153 . 24. Pyron RA , Burbrink FT , Colli GR , Montes de Oca AN , Vitt LJ , Kuczynski CA , et al. The phylogeny of advanced snakes (Colubroidea), with discovery of a new subfamily and comparison of support methods for likelihood trees . Mol Phylogenet Evol . 2011 ; 58 : 329 - 342 . https://doi.org/10.1016/j.ympev. 2010 . 11 .006 PMID: 21074626 25. Zaher H , Grazziotin FG , Graboski R , Fuentes RG , Sa?nchez- Martinez P , Montingelli GG , et al. Phylogenetic relationships of the genus Sibynophis (Serpentes: Colubroidea) . Pap Avulsos Zool . 2012 ; 52 : 141 - 149 . 26. Pyron RA , Burbrink FT , Wiens JJ . A phylogeny and revised classification of Squamata, including 4161 species of lizards and snakes . BMC Evol Biol . 2013 ; 13 : 93 . https://doi.org/10.1186/ 1471 -2148-13-93 PMID: 23627680 27. Zheng Y , Wiens JJ . Combining phylogenomic and supermatrix approaches, and a time-calibrated phylogeny for squamate reptiles (lizards and snakes) based on 52 genes and 4162 species . Mol. Phyl. Evol . 2016 ; 94 : 537 - 547 . 28. Figueroa A , McKelvy AD , Grismer LL , Bell CD , Lailvaux SP . A species-level phylogeny of extant snakes with description of a new colubrid subfamily and genus . PLoS One . 2016 ; 11 : e0161070. https://doi.org/10.1371/journal.pone. 0161070 PMID: 27603205 29. Zaher H . Hemipenial morphology of the South American xenodontine snakes, with a proposal for a monophyletic Xenodontinae and a reappraisal of colubroid hemipenes . Bull Am Mus Nat Hist . 1999 ; 240 : 1 - 168 . 30. Cundall D , Irish FJ . The snake skull . In: Gans C , Gaunt AS , Adler K , editors. Biology of the Reptilia , vol 20 . Ithaca: Society for the Study of Amphibians and Reptiles; 2008 . pp. 349 - 692 . 31. Rage JC. Serpentes . In: Wellnhofer P, editor. Handbuch der Palaoherpetologie. Stuttgart: Gustav Fischer; 1984 . pp. 1 - 80 . 32. Rage JC . Fossil history . In: Seigel RA , Collins JT , Novak SS , editors. Snakes, ecology and evolutionary biology . New York: Macmillan Publishing Company; 1987 . pp. 51 - 76 . 33. Uetz P , Freed P , Hos?ek J . The Reptile Database; 2017 [cited 2018 Jan 10 ]. http://www.reptiledatabase.org. 45. Phillippe H , Snell EA , Bapteste E , Lopez P , Holland PWH , Casane D. Phylogenomics of eukaryotes: impact of missing data on large alignments . Mol Biol Evol . 2004 ; 21 : 1740 - 1752 . https://doi.org/10. 1093/molbev/msh182 PMID: 15175415 46. Lanfear R. , Calcott B. , Ho S.Y. , Guindon S. PartitionFinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses . Mol. Biol. Evol . 2012 ; 29 : 1695 - 1701 . https://doi.org/10.1093/molbev/mss020 PMID: 22319168 48. Felsenstein J . Confidence limits on phylogenies: an approach using the bootstrap . Evolution . 1985 ; 39 : 783 - 791 . https://doi.org/10.1111/j.1558- 5646 . 1985 .tb00420. x PMID: 28561359 49. Lemoine F , Entfellner JBD , Wilkinson E , Correia D , Felipe MD , Oliveira T , Gascuel O. Renewing Felsenstein's phylogenetic bootstrap in the era of big data . Nature . 2018 ; 556 : 452 - 456 . https://doi.org/ 10.1038/s41586-018 -0043-0 PMID: 29670290 50. Kozlov A. , Darriba D. , Flouri T. , Morel B. , Stamatakis A . RAxML-NG : A fast, scalable, and user-friendly tool for maximum likelihood phylogenetic inference . bioRxiv . 2018 : 447110 . https://doi.org/10.1101/ 51. Guindon S , Dufayard JF , Lefort V , Anisimova M , Hordijk W , Gascuel O . New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0 . Syst . Biol. 2010 ; 59 : 307 - 321 . https://doi.org/10.1093/sysbio/syq010 PMID: 20525638 52. Drummond AJ , Suchard MA , Xie D , Rambaut A . Bayesian phylogenetics with BEAUti and the BEAST 1.7 . Mol Biol Evol . 2012 ; 29 : 1969 - 1973 . https://doi.org/10.1093/molbev/mss075 PMID: 22367748 69. Head JJ , Mahlow K , Mu?ller J. Fossil calibration dates for molecular phylogenetic analysis of snakes 2: Caenophidia , Colubroidea, Elapoidea, Colubridae. Palaeo Electronica . 2016 ; 19 : 1 - 21 . 70. Rana RS , Auge? M, Folie A , Rose KD , Kumar K , Singh L , Sahni A , Smith T. High diversity of acrodontan lizards in the Early Eocene Vastan Lignite Mine of India . Geol Belg . 2013 ; 16 : 290 - 301 . 71. Sahni A , Saraswati PK , Rana RS , Kumar K , Singh H , Alimohammadian H , Sahni N , Rose KD , Singh L , Smith T. Temporal constraints and depositional palaeoenvironments of the Vastan lignite sequence, Gujarat: analogy for the Cambay Shale hydrocarbon source rock . Indian J Petrol Geol . 2006 ; 15 : 1 - 20 . 72. Rana RS , Kumar K , Escarguel G , Sahni A , Rose KD , Smith T , Singh H , Singh L . An ailuravine rodent from the lower Eocene Cambay Formation at Vastan, western India, and its palaeobiogeographic implications . Acta Palaeontol Polonica . 2008 ; 53 : 1 - 14 . 73. Schaub H. Nummulites et Assilines de la Te? thys pale?og?ne. Taxinomie, phylogen?se et biostratigraphie . Me?m Suisses Pale?ontol. 1981 ; 104 : 1 - 236 . 74. Serra-Kiel J , Hottinger L , Caus E , Drobne K , Ferrandez C , Jauhri AK , et al. Larger foraminiferal biostratigraphy of the Tethyan Paleocene and Eocene . Bull Soc Ge?ol France . 1998 ; 169 : 281 - 299 . 75. Vandenberghe N , Hilgen FJ , Speijer RP . The Paleogene Period . In: Gradstein F, Ogg J , Schmitz M , Ogg G., editors. The Geologic Time Scale 2012 . London: Elsevier; 2012 . pp. 855 - 921 . 76. Garg R , Khowaja A , Prasad V , Tripathi SKM , Singh IB , Jauhri AK , et al. Age-diagnostic dinoflagellate cysts from lignite-bearing sediments of the Vastan lignite mine , Surat District, Gujarat, western India. J Palaeontol Soc India . 2008 ; 53 : 99 - 105 . 92. Agust ?? J. A criticial reevaluation of the Miocene mammal units in Western Europe: Dispersal events and problems of correlation . In Augusti J , Rook L , Andrews P , editors, The evolution of Neogene terrestrial ecosystems in Europe . Cambridge: Cambridge University; 1999 . pp. 84 - 112 . 93. Vangengeim E , Tesakov A . Late Miocene mammal localities of eastern Europe and western Asia . In: Wang X , Flynn LJ , Fortelius M , editors. Fossil mammals of Asia . Neogene biostratigraphy and chronology . New York, Columbia University Press; 2013 . pp. 521 - 545 . 94. McCartney JM , Stevens NJ , O'Connor PM. The earliest colubroid-dominated snake fauna from Africa: Perspectives from the late Oligocene Nsungwe Formation of southwestern Tanzania . PLoS ONE . 2014 ; 9: e90415 . https://doi.org/10.1371/journal.pone. 0090415 PMID: 24646522 95. Stevens NJ , Ngasala S , Gottfried MD , O'Connor PM , Roberts EM . Macroscelideans from the Oligocene of southwestern Tanzania . J Vert Paleontol . 2006 ; 26 : 128A . 96. Stevens NJ , O'Connor PM , Gottfried MD , Roberts EM , Ngasala S , et al. Metaphiomys (Rodentia: Phiomyidae) from the Paleogene of southwestern Tanzania . J Paleontol . 2006 ; 80 : 407 - 410 . 97. Stevens NJ , Gottfried MD , Roberts EM , Kapilima S , Ngasala S , et al. Paleontological exploration in Africa: a view from the Rukwa Rift Basin of Tanzania . In: Fleagle JG , Gilbert CC , editors. Elwyn Simons: A Search for Origins . New York: Springer; 2008 . pp. 159 - 180 . 115. Szyndlar Z. A review of Neogene and Quaternary snakes of central and eastern Europe. Part II . Natricinae, Elapidae, Viperidae. Estud Geol-Madrid. 1991 ; 47 : 237 - 266 . 141. Zaher H , Scanferla CA. The skull of the Upper Cretaceous snake Dinilysia patagonica Smith- Woodward , 1901 , and its phylogenetic . Zool J Linn Soc . 2012 ; 164 : 194 - 238 . Rage JC . Les quamates du Mioc?ne de Beni Mellal, Maroc. Geol Medit . 1976 ; 3 : 57 - 70 . 190. Szyndlar Z , Zerova G . Neogene cobras of the genus Naja (Serpentes: Elapidae) of East Europe . Ann Naturhist Mus Wien. 1990 ; 91 : 53 - 61 . 191. Szyndlar Z , Rage JC . West Palearctic cobras of the genus Naja (Serpentes: Elapidae): interrelationships among extinct and extant species . Amphibia-Reptilia . 1990 ; 11 : 385 - 400 . 192. Bachmayer F , Szyndlar Z. Ophidians (Reptilia: Serpentes) from the Kohfidisch fissures of Burgenland, Austria . Ann Naturhist Mus Wien Ser A. 1985 ; 87 : 79 - 100 . 193. Rage JC , Holman JA . Des serpents (Reptilia, Squamata) de type nord-ame?ricain dans le Mioc?ne franc?ais. E?volution parall?le ou dispersion? Geobios . 1984 ; 17 : 89 - 104 . 194. Scanlon JD , Lee MSY . Phylogeny of Australasian venomous snakes (Colubroidea, Elapidae, Hydrophiinae) based on phenotypic and molecular evidence . Zool Scr . 2004 ; 33 : 335 - 366 . 195. Holman JA . Upper Miocene snakes (Reptilia, Serpentes) from southeastern Nebraska . J Herpetol . 1977 ; 11 : 323 - 335 . 196. Myers CW. A new genus and new tribe for Enicognathus melanauchen Jan, 1863, a neglected South American snake (Colubridae: Xenodontinae), with taxonomic notes on some Dipsadinae . Amer Mus Novit . 2011 ; 3715 : 1 - 33 . 197. Parham JF , Donoghue PCJ , Bell CJ , Calways TD , Head JJ , Holroyd PA et al. Best Practices for Justifying Fossil Calibrations . Syst. Biol . 2012 ; 61 : 346 - 359 . https://doi.org/10.1093/sysbio/syr107 PMID: 22105867


This is a preview of a remote PDF: https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0216148&type=printable

Hussam Zaher, Robert W. Murphy, Juan Camilo Arredondo, Roberta Graboski, Paulo Roberto Machado-Filho, Kristin Mahlow, Giovanna G. Montingelli, Ana Bottallo Quadros, Nikolai L. Orlov, Mark Wilkinson, Ya-Ping Zhang, Felipe G. Grazziotin. Large-scale molecular phylogeny, morphology, divergence-time estimation, and the fossil record of advanced caenophidian snakes (Squamata: Serpentes), PLOS ONE, 2019, DOI: 10.1371/journal.pone.0216148